open-swe/tests/test_plan_mode.py

227 lines
7.2 KiB
Python
Raw Normal View History

feat: plan mode with model-driven entry and collaborative review (#1580) * feat: add plan mode for read-only research and planning Adds a per-run plan_mode flag that puts the agent in a read-only research phase: a strong prompt section is injected and mutating tools are stripped via ExcludeToolsMiddleware so the agent proposes a reviewable implementation plan before any edits. Surfaced in the dashboard UI with a Plan toggle (Shift+Tab) wired through the thread API. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: enforce plan-mode read-only at tool layer and disable subagents Addresses PR review: plan mode previously relied on prompt text to keep the shell read-only and left the task subagent (built with its own write/PR/Linear tools) unrestricted. Now `task` is excluded so research cannot be delegated to a mutating subagent, and a new PlanModeShellGuardMiddleware enforces a read-only command allowlist on `execute`, blocking writes, git state changes, installs, redirection, and command substitution regardless of model/prompt-injection compliance. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: harden plan-mode shell guard against wrapped mutations Block git global options that take values (-C, --git-dir, ...) from being misread as the subcommand, reject config-injection options (-c, --config-env, --exec-path), and drop the env command wrapper that could run arbitrary commands. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * feat: add plan mode with enter_plan_mode tool, profile/team defaults, Slack commands and approval flow - enter_plan_mode tool: agent self-activates plan mode via Command(update={'plan_mode': True}) - Plan mode resolution: per-thread > profile default > team default > False - PLAN_MODE_GUIDANCE_SECTION: always-present prompt section telling agent about the tool - profile_plan_mode_default and team plan_mode_default settings - Slack plan on/off/status commands with thread metadata persistence - slack_thread_reply plan_approval=True renders Approve/Revise/Cancel buttons - Interactivity handler: approve triggers implementation run, cancel posts confirmation - Frontend: plan_mode_default in Profile/ProfileUpdate/TeamSettings types and UI toggles Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * test: add tests for enter_plan_mode tool, profile/team defaults, Slack plan commands, approval blocks Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * refactor(plan-mode): drop shell guard, rely on prompt for read-only discipline Remove PlanModeShellGuardMiddleware and its enforcement of read-only shell commands during plan mode. Plan mode now relies on the system prompt to instruct the agent not to run mutating commands; the mutating-tool exclusion (ExcludeToolsMiddleware) is retained. * test(open-swe): add Playwright E2E for the Slack → PR → web handoff Local, secrets-free end-to-end suite that drives the full happy path through mock Slack/GitHub control panels and the real dashboard UI. Only the LLM and external SaaS HTTP boundaries (GitHub/Slack APIs, OAuth token mint) are faked — the real process_slack_mention, get_agent, deepagents loop, tools, middleware, and dashboard authorization all run under `langgraph dev` with a scripted fake chat model and a local temp-dir sandbox. - full_flow: a Slack mention runs the agent, which implements a change in the sandbox, opens a PR against a fake GitHub remote, and replies with the PR link in the same thread. - dashboard: clicking the bot's real "Open in Web" link loads the built ui/ app (served same-origin); the thread owner can continue the conversation, while a different user sees the same thread read-only (no composer). Wired into Agent CI as a `Playwright E2E` job that runs on pull requests. * fix(open-swe): serve E2E UI assets via explicit route; pin Playwright The dashboard E2E served the built ui/ SPA's /assets via app.mount(StaticFiles), but LangGraph's custom-app loader serves APIRoutes and drops sub-app Mounts, so /assets 404'd under `langgraph dev` in CI — the React app never booted and the composer/transcript never rendered. Serve assets via an explicit route instead. Also pin @playwright/test to the latest (1.61.0) for reproducible runs, and make the owner composer assertion tolerant of either hydration state. * test(open-swe): record Playwright trace + video on every E2E run Capture a replayable trace (DOM snapshots, network, console, source) and a screen recording for every test, not just retries, plus a screenshot on failure. The CI job already uploads playwright-report/ and test-results/, so each run now has a downloadable replay; documented how to open it. * feat(plan-mode): collaborative plan review with BlockNote + Yjs When the agent enters plan mode it writes the plan as a markdown file in the sandbox (save_plan tool), publishes it, and posts a review link to the source channel. Reviewers open the plan inside the dashboard (under the /agents shell), read it rendered in a BlockNote editor, and leave inline comments synced live over Yjs. Only the thread owner can approve; any reviewer can request changes. On approve/reject the comments are harvested and handed to the agent for the follow-up run; the agent never sees comments mid-review. - agent: enter_plan_mode persists plan state; new save_plan tool; prompt shares the plan-review link. - dashboard: Yjs WebSocket collab server (pycrdt-websocket) with store-backed snapshots; plan content/status store; plan REST API (get/approve/reject, owner-only approve, client-harvested comments); planStatus on thread summaries. - ui: BlockNote native comments (CommentsExtension + YjsThreadStore) plan page mounted under the agents shell, with a "Review plan" banner in the thread view and a back-link; theme-aware (dark mode) using the dashboard tokens. - e2e: Playwright coverage of the full Slack -> plan -> review -> approve -> PR flow, including cross-user comment sync and owner-only approval. * fix(plan-mode): address review feedback (authz, overrides, leaks, deps) - plan-collab WS: authorize per-thread before joining a room (same read gate as the REST API) — previously any logged-in user could join any thread (IDOR). - plan-collab: tie the snapshot flusher to active connections (refcount) so each opened plan no longer leaks a permanent 1.5s task on the shared event loop. - plan decisions: include thread_id in the follow-up run configurable so the run resumes the existing thread; set plan_mode explicitly so approve forces it off. - get_agent: an explicit per-thread plan_mode (Slack `plan off`, approved plan, dashboard toggle) now overrides profile/team defaults instead of falling back. - plan mode tool gating moved to a state-aware PlanModeMiddleware installed unconditionally, so a mid-run enter_plan_mode restricts the next model turn; before_agent resets stale plan_mode so a later run isn't forced back into it. - exclude write-capable http_request from plan mode. - pin pycrdt / pycrdt-websocket with upper bounds. Includes the latest base (#1583): E2E UI assets served via explicit route (fixes the Playwright CI failure — LangGraph's app loader drops sub-app mounts). * style: ruff format plan_collab.py * fix(plan-mode): owner-gate Slack approval + same-origin check on collab WS - Slack "Approve & Implement" now verifies the clicking user is the plan requester (owner, via the stored triggering_user_id) before implementing — matching the dashboard API's owner-only approval. Non-owners are pointed to Revise / feedback. - The plan-collab WebSocket validates the handshake Origin against the dashboard allowlist before accept() (no-op when unconfigured, e.g. local/dev), mirroring the REST require_same_origin CSRF defense. * fix(plan-mode): enter plan mode only via the model + local mock dev harness Plan mode is now entered solely when the model calls enter_plan_mode. Removed the per-user and team plan_mode_default settings (backend + UI) and the Slack `plan on/off/status` toggle. - enter_plan_mode returns a terminating ToolMessage, fixing the missing ToolMessage error that silently dropped plan mode mid-run. - PlanReview: defer Yjs provider/doc teardown so React StrictMode's dev remount doesn't destroy and then reuse the collaboration provider. - e2e plan_review spec asserts plan_mode actually engages. - LangSmith trace-url resolution is best-effort: bail before any API call when the tenant is unset, cache failures, log at debug. - Add `pnpm run dev:mock`: same-origin Vite HMR harness with a real LLM, Alice/Bob mock users, and a GitHub login picker. * docs(plan-mode): drop stale references to removed profile/team defaults The plan_mode middleware docstring and the approve/reject dispatch comment still described the profile/team plan_mode_default resolution that no longer exists; reword to match model-driven entry + the per-thread carry. * feat(plan-mode): let any reviewer edit the plan, not just comment Drop the owner/commenter split for the plan document: everyone with read access edits and comments alike (DefaultThreadStoreAuth "editor" for all, editor always editable until a decision, anyone seeds the empty doc). This matches the collab WS, which already relays frames to every readable user. Plan approval stays owner-gated. * test(plan-mode): assert plan-mode entry via the tool's success message plan_mode lives only in run state for tool gating; it is not a persisted thread-state channel, so the previous `values.plan_mode === true` poll could never pass. Assert instead that enter_plan_mode's success ToolMessage ("Plan mode is active …") lands in the thread — which only happens when the tool's Command applies cleanly, the exact regression this guards. --------- Co-authored-by: Johannes du Plessis <51395795+johannes117@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
2026-06-23 15:06:58 -04:00
from __future__ import annotations
import asyncio
from typing import Any
import pytest
from agent import server
from agent.dashboard import thread_api
from agent.prompt import construct_system_prompt
def test_plan_mode_prompt_included_when_enabled() -> None:
prompt = construct_system_prompt(working_dir="/work", plan_mode=True)
assert "Plan Mode (ACTIVE)" in prompt
assert "read-only research-and-planning phase" in prompt
def test_plan_mode_prompt_absent_by_default() -> None:
prompt = construct_system_prompt(working_dir="/work")
assert "Plan Mode (ACTIVE)" not in prompt
def test_plan_mode_excluded_tools_cover_mutating_tools() -> None:
excluded = server.PLAN_MODE_EXCLUDED_TOOLS
for tool in (
"task",
"open_pull_request",
"request_pr_review",
"slack_start_new_thread",
feat: plan mode with model-driven entry and collaborative review (#1580) * feat: add plan mode for read-only research and planning Adds a per-run plan_mode flag that puts the agent in a read-only research phase: a strong prompt section is injected and mutating tools are stripped via ExcludeToolsMiddleware so the agent proposes a reviewable implementation plan before any edits. Surfaced in the dashboard UI with a Plan toggle (Shift+Tab) wired through the thread API. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: enforce plan-mode read-only at tool layer and disable subagents Addresses PR review: plan mode previously relied on prompt text to keep the shell read-only and left the task subagent (built with its own write/PR/Linear tools) unrestricted. Now `task` is excluded so research cannot be delegated to a mutating subagent, and a new PlanModeShellGuardMiddleware enforces a read-only command allowlist on `execute`, blocking writes, git state changes, installs, redirection, and command substitution regardless of model/prompt-injection compliance. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: harden plan-mode shell guard against wrapped mutations Block git global options that take values (-C, --git-dir, ...) from being misread as the subcommand, reject config-injection options (-c, --config-env, --exec-path), and drop the env command wrapper that could run arbitrary commands. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * feat: add plan mode with enter_plan_mode tool, profile/team defaults, Slack commands and approval flow - enter_plan_mode tool: agent self-activates plan mode via Command(update={'plan_mode': True}) - Plan mode resolution: per-thread > profile default > team default > False - PLAN_MODE_GUIDANCE_SECTION: always-present prompt section telling agent about the tool - profile_plan_mode_default and team plan_mode_default settings - Slack plan on/off/status commands with thread metadata persistence - slack_thread_reply plan_approval=True renders Approve/Revise/Cancel buttons - Interactivity handler: approve triggers implementation run, cancel posts confirmation - Frontend: plan_mode_default in Profile/ProfileUpdate/TeamSettings types and UI toggles Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * test: add tests for enter_plan_mode tool, profile/team defaults, Slack plan commands, approval blocks Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * refactor(plan-mode): drop shell guard, rely on prompt for read-only discipline Remove PlanModeShellGuardMiddleware and its enforcement of read-only shell commands during plan mode. Plan mode now relies on the system prompt to instruct the agent not to run mutating commands; the mutating-tool exclusion (ExcludeToolsMiddleware) is retained. * test(open-swe): add Playwright E2E for the Slack → PR → web handoff Local, secrets-free end-to-end suite that drives the full happy path through mock Slack/GitHub control panels and the real dashboard UI. Only the LLM and external SaaS HTTP boundaries (GitHub/Slack APIs, OAuth token mint) are faked — the real process_slack_mention, get_agent, deepagents loop, tools, middleware, and dashboard authorization all run under `langgraph dev` with a scripted fake chat model and a local temp-dir sandbox. - full_flow: a Slack mention runs the agent, which implements a change in the sandbox, opens a PR against a fake GitHub remote, and replies with the PR link in the same thread. - dashboard: clicking the bot's real "Open in Web" link loads the built ui/ app (served same-origin); the thread owner can continue the conversation, while a different user sees the same thread read-only (no composer). Wired into Agent CI as a `Playwright E2E` job that runs on pull requests. * fix(open-swe): serve E2E UI assets via explicit route; pin Playwright The dashboard E2E served the built ui/ SPA's /assets via app.mount(StaticFiles), but LangGraph's custom-app loader serves APIRoutes and drops sub-app Mounts, so /assets 404'd under `langgraph dev` in CI — the React app never booted and the composer/transcript never rendered. Serve assets via an explicit route instead. Also pin @playwright/test to the latest (1.61.0) for reproducible runs, and make the owner composer assertion tolerant of either hydration state. * test(open-swe): record Playwright trace + video on every E2E run Capture a replayable trace (DOM snapshots, network, console, source) and a screen recording for every test, not just retries, plus a screenshot on failure. The CI job already uploads playwright-report/ and test-results/, so each run now has a downloadable replay; documented how to open it. * feat(plan-mode): collaborative plan review with BlockNote + Yjs When the agent enters plan mode it writes the plan as a markdown file in the sandbox (save_plan tool), publishes it, and posts a review link to the source channel. Reviewers open the plan inside the dashboard (under the /agents shell), read it rendered in a BlockNote editor, and leave inline comments synced live over Yjs. Only the thread owner can approve; any reviewer can request changes. On approve/reject the comments are harvested and handed to the agent for the follow-up run; the agent never sees comments mid-review. - agent: enter_plan_mode persists plan state; new save_plan tool; prompt shares the plan-review link. - dashboard: Yjs WebSocket collab server (pycrdt-websocket) with store-backed snapshots; plan content/status store; plan REST API (get/approve/reject, owner-only approve, client-harvested comments); planStatus on thread summaries. - ui: BlockNote native comments (CommentsExtension + YjsThreadStore) plan page mounted under the agents shell, with a "Review plan" banner in the thread view and a back-link; theme-aware (dark mode) using the dashboard tokens. - e2e: Playwright coverage of the full Slack -> plan -> review -> approve -> PR flow, including cross-user comment sync and owner-only approval. * fix(plan-mode): address review feedback (authz, overrides, leaks, deps) - plan-collab WS: authorize per-thread before joining a room (same read gate as the REST API) — previously any logged-in user could join any thread (IDOR). - plan-collab: tie the snapshot flusher to active connections (refcount) so each opened plan no longer leaks a permanent 1.5s task on the shared event loop. - plan decisions: include thread_id in the follow-up run configurable so the run resumes the existing thread; set plan_mode explicitly so approve forces it off. - get_agent: an explicit per-thread plan_mode (Slack `plan off`, approved plan, dashboard toggle) now overrides profile/team defaults instead of falling back. - plan mode tool gating moved to a state-aware PlanModeMiddleware installed unconditionally, so a mid-run enter_plan_mode restricts the next model turn; before_agent resets stale plan_mode so a later run isn't forced back into it. - exclude write-capable http_request from plan mode. - pin pycrdt / pycrdt-websocket with upper bounds. Includes the latest base (#1583): E2E UI assets served via explicit route (fixes the Playwright CI failure — LangGraph's app loader drops sub-app mounts). * style: ruff format plan_collab.py * fix(plan-mode): owner-gate Slack approval + same-origin check on collab WS - Slack "Approve & Implement" now verifies the clicking user is the plan requester (owner, via the stored triggering_user_id) before implementing — matching the dashboard API's owner-only approval. Non-owners are pointed to Revise / feedback. - The plan-collab WebSocket validates the handshake Origin against the dashboard allowlist before accept() (no-op when unconfigured, e.g. local/dev), mirroring the REST require_same_origin CSRF defense. * fix(plan-mode): enter plan mode only via the model + local mock dev harness Plan mode is now entered solely when the model calls enter_plan_mode. Removed the per-user and team plan_mode_default settings (backend + UI) and the Slack `plan on/off/status` toggle. - enter_plan_mode returns a terminating ToolMessage, fixing the missing ToolMessage error that silently dropped plan mode mid-run. - PlanReview: defer Yjs provider/doc teardown so React StrictMode's dev remount doesn't destroy and then reuse the collaboration provider. - e2e plan_review spec asserts plan_mode actually engages. - LangSmith trace-url resolution is best-effort: bail before any API call when the tenant is unset, cache failures, log at debug. - Add `pnpm run dev:mock`: same-origin Vite HMR harness with a real LLM, Alice/Bob mock users, and a GitHub login picker. * docs(plan-mode): drop stale references to removed profile/team defaults The plan_mode middleware docstring and the approve/reject dispatch comment still described the profile/team plan_mode_default resolution that no longer exists; reword to match model-driven entry + the per-thread carry. * feat(plan-mode): let any reviewer edit the plan, not just comment Drop the owner/commenter split for the plan document: everyone with read access edits and comments alike (DefaultThreadStoreAuth "editor" for all, editor always editable until a decision, anyone seeds the empty doc). This matches the collab WS, which already relays frames to every readable user. Plan approval stays owner-gated. * test(plan-mode): assert plan-mode entry via the tool's success message plan_mode lives only in run state for tool gating; it is not a persisted thread-state channel, so the previous `values.plan_mode === true` poll could never pass. Assert instead that enter_plan_mode's success ToolMessage ("Plan mode is active …") lands in the thread — which only happens when the tool's Command applies cleanly, the exact regression this guards. --------- Co-authored-by: Johannes du Plessis <51395795+johannes117@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
2026-06-23 15:06:58 -04:00
"linear_create_issue",
"linear_update_issue",
"linear_delete_issue",
):
assert tool in excluded
# Read-only tools must stay available.
assert "read_file" not in excluded
assert "execute" not in excluded
feat: Re-land deferred upstream features on modular webhooks (#80) (#128) * fix(webhooks): fall back to vision model for Slack/Linear image threads Re-land upstream #1626 onto the modular webhook structure. When a Slack mention or Linear issue carries images but the resolved model is text-only, fall back to a vision-capable model instead of dropping the images. Re-points default_vision_model_pair at the fork's image-capable models (Opus 4.8 default, else any supports_images model) rather than upstream's openai:/anthropic: provider filter. Refs #80, upstream #1626 * fix(slack): persist trace_message_ts so web-handoff updates the trace reply Re-land upstream #1630 onto the modular structure. The first-mention store_slack_run_mapping call did not pass trace_message_ts, so it was never persisted (nothing to preserve from on first mention) and _notify_slack_web_handoff always skipped the trace-reply update on web handoff. Pass it through and cover it with a test. Refs #80, upstream #1630 * feat(slack): include channel context in Slack prompts Re-land upstream #1633 onto the modular structure. Fetch cached Slack channel metadata once per event (_get_slack_channel_context) and thread it through the docs-plz gate, repo resolution, and process_slack_mention so prompts carry the channel name and a clearly-marked untrusted channel description. Avoids duplicate conversations.info calls. Refs #80, upstream #1633 * feat(tools): add slack_start_new_thread breakout tool Re-land upstream #1638 onto the modular structure. Adds the slack_start_new_thread tool (posts a top-level Slack message and dispatches a fresh agent run for a broken-out task via the durable dispatch_agent_run contract), wires it into the agent tool list and tools/__init__, adds prompt guidance, and excludes it from plan mode so it can't bypass the approval flow. Tool imports only live modules. Refs #80, upstream #1638 * feat(plan): notify Slack on plan approval Re-land upstream #1632 onto the modular structure. When a plan is approved via the dashboard approve endpoint, post a thread reply to the originating Slack thread noting the comment count and approver, after the follow-up run is dispatched. Slack post failures never break approval. Adapted to the fork's approve_plan (no plan_markdown read). Refs #80, upstream #1632 * feat(plan): publish plans from sandbox files Re-land upstream #1635 onto the modular structure, completing the partially-ported change so dev is internally consistent. save_plan now takes a plan_file_path, reads the agent-authored Markdown file from /workspace/plans/ (validating extension/location/UTF-8/size) and publishes it, instead of taking a plan_markdown string. Removes write_file/edit_file from PLAN_MODE_EXCLUDED_TOOLS so the agent can author the plan file, updates enter_plan_mode/reject_plan guidance and the e2e fake LLM. Skips the #1610-only update_plan hunk (not on dev). Refs #80, upstream #1635 * fix(security): SSRF-harden server-side image fetch + stop logging raw image URLs INJ-01 (high): fetch_image_block used follow_redirects=True with no per-hop revalidation and discarded the resolved-IP pin, so an attacker-authored Slack/ Linear image URL could 302-redirect the fetch to an internal host / cloud metadata endpoint (blind SSRF), and DNS-rebinding could bypass the one-shot is_url_safe check. Route image fetches through the same per-hop resolve+pin+ revalidate loop the http_request tool uses, lifted into url_safety as the shared request_with_safe_redirects. Also strip the per-host Slack/Linear bearer token on redirect so it can't be replayed to a redirect target. SC-1 (low): linear.py logged full image URLs (which can carry signed tokens) at DEBUG; multimodal logged them at INFO on every fetch. Log host-only. Sink lived in multimodal.py (unchanged by the feature work) but PR #128 widened its reach by no longer dropping images for text-only models. Fixing on the base branch so #130/#129 inherit it on rebase. Adds fetch_image_block SSRF regression tests (redirect-to-internal blocked; auth stripped on redirect).
2026-07-08 18:32:43 -04:00
# File edit tools stay available so the agent can author the plan file under
# /workspace/plans/ (restricted to that path via prompt guidance).
assert "write_file" not in excluded
assert "edit_file" not in excluded
feat: plan mode with model-driven entry and collaborative review (#1580) * feat: add plan mode for read-only research and planning Adds a per-run plan_mode flag that puts the agent in a read-only research phase: a strong prompt section is injected and mutating tools are stripped via ExcludeToolsMiddleware so the agent proposes a reviewable implementation plan before any edits. Surfaced in the dashboard UI with a Plan toggle (Shift+Tab) wired through the thread API. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: enforce plan-mode read-only at tool layer and disable subagents Addresses PR review: plan mode previously relied on prompt text to keep the shell read-only and left the task subagent (built with its own write/PR/Linear tools) unrestricted. Now `task` is excluded so research cannot be delegated to a mutating subagent, and a new PlanModeShellGuardMiddleware enforces a read-only command allowlist on `execute`, blocking writes, git state changes, installs, redirection, and command substitution regardless of model/prompt-injection compliance. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: harden plan-mode shell guard against wrapped mutations Block git global options that take values (-C, --git-dir, ...) from being misread as the subcommand, reject config-injection options (-c, --config-env, --exec-path), and drop the env command wrapper that could run arbitrary commands. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * feat: add plan mode with enter_plan_mode tool, profile/team defaults, Slack commands and approval flow - enter_plan_mode tool: agent self-activates plan mode via Command(update={'plan_mode': True}) - Plan mode resolution: per-thread > profile default > team default > False - PLAN_MODE_GUIDANCE_SECTION: always-present prompt section telling agent about the tool - profile_plan_mode_default and team plan_mode_default settings - Slack plan on/off/status commands with thread metadata persistence - slack_thread_reply plan_approval=True renders Approve/Revise/Cancel buttons - Interactivity handler: approve triggers implementation run, cancel posts confirmation - Frontend: plan_mode_default in Profile/ProfileUpdate/TeamSettings types and UI toggles Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * test: add tests for enter_plan_mode tool, profile/team defaults, Slack plan commands, approval blocks Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * refactor(plan-mode): drop shell guard, rely on prompt for read-only discipline Remove PlanModeShellGuardMiddleware and its enforcement of read-only shell commands during plan mode. Plan mode now relies on the system prompt to instruct the agent not to run mutating commands; the mutating-tool exclusion (ExcludeToolsMiddleware) is retained. * test(open-swe): add Playwright E2E for the Slack → PR → web handoff Local, secrets-free end-to-end suite that drives the full happy path through mock Slack/GitHub control panels and the real dashboard UI. Only the LLM and external SaaS HTTP boundaries (GitHub/Slack APIs, OAuth token mint) are faked — the real process_slack_mention, get_agent, deepagents loop, tools, middleware, and dashboard authorization all run under `langgraph dev` with a scripted fake chat model and a local temp-dir sandbox. - full_flow: a Slack mention runs the agent, which implements a change in the sandbox, opens a PR against a fake GitHub remote, and replies with the PR link in the same thread. - dashboard: clicking the bot's real "Open in Web" link loads the built ui/ app (served same-origin); the thread owner can continue the conversation, while a different user sees the same thread read-only (no composer). Wired into Agent CI as a `Playwright E2E` job that runs on pull requests. * fix(open-swe): serve E2E UI assets via explicit route; pin Playwright The dashboard E2E served the built ui/ SPA's /assets via app.mount(StaticFiles), but LangGraph's custom-app loader serves APIRoutes and drops sub-app Mounts, so /assets 404'd under `langgraph dev` in CI — the React app never booted and the composer/transcript never rendered. Serve assets via an explicit route instead. Also pin @playwright/test to the latest (1.61.0) for reproducible runs, and make the owner composer assertion tolerant of either hydration state. * test(open-swe): record Playwright trace + video on every E2E run Capture a replayable trace (DOM snapshots, network, console, source) and a screen recording for every test, not just retries, plus a screenshot on failure. The CI job already uploads playwright-report/ and test-results/, so each run now has a downloadable replay; documented how to open it. * feat(plan-mode): collaborative plan review with BlockNote + Yjs When the agent enters plan mode it writes the plan as a markdown file in the sandbox (save_plan tool), publishes it, and posts a review link to the source channel. Reviewers open the plan inside the dashboard (under the /agents shell), read it rendered in a BlockNote editor, and leave inline comments synced live over Yjs. Only the thread owner can approve; any reviewer can request changes. On approve/reject the comments are harvested and handed to the agent for the follow-up run; the agent never sees comments mid-review. - agent: enter_plan_mode persists plan state; new save_plan tool; prompt shares the plan-review link. - dashboard: Yjs WebSocket collab server (pycrdt-websocket) with store-backed snapshots; plan content/status store; plan REST API (get/approve/reject, owner-only approve, client-harvested comments); planStatus on thread summaries. - ui: BlockNote native comments (CommentsExtension + YjsThreadStore) plan page mounted under the agents shell, with a "Review plan" banner in the thread view and a back-link; theme-aware (dark mode) using the dashboard tokens. - e2e: Playwright coverage of the full Slack -> plan -> review -> approve -> PR flow, including cross-user comment sync and owner-only approval. * fix(plan-mode): address review feedback (authz, overrides, leaks, deps) - plan-collab WS: authorize per-thread before joining a room (same read gate as the REST API) — previously any logged-in user could join any thread (IDOR). - plan-collab: tie the snapshot flusher to active connections (refcount) so each opened plan no longer leaks a permanent 1.5s task on the shared event loop. - plan decisions: include thread_id in the follow-up run configurable so the run resumes the existing thread; set plan_mode explicitly so approve forces it off. - get_agent: an explicit per-thread plan_mode (Slack `plan off`, approved plan, dashboard toggle) now overrides profile/team defaults instead of falling back. - plan mode tool gating moved to a state-aware PlanModeMiddleware installed unconditionally, so a mid-run enter_plan_mode restricts the next model turn; before_agent resets stale plan_mode so a later run isn't forced back into it. - exclude write-capable http_request from plan mode. - pin pycrdt / pycrdt-websocket with upper bounds. Includes the latest base (#1583): E2E UI assets served via explicit route (fixes the Playwright CI failure — LangGraph's app loader drops sub-app mounts). * style: ruff format plan_collab.py * fix(plan-mode): owner-gate Slack approval + same-origin check on collab WS - Slack "Approve & Implement" now verifies the clicking user is the plan requester (owner, via the stored triggering_user_id) before implementing — matching the dashboard API's owner-only approval. Non-owners are pointed to Revise / feedback. - The plan-collab WebSocket validates the handshake Origin against the dashboard allowlist before accept() (no-op when unconfigured, e.g. local/dev), mirroring the REST require_same_origin CSRF defense. * fix(plan-mode): enter plan mode only via the model + local mock dev harness Plan mode is now entered solely when the model calls enter_plan_mode. Removed the per-user and team plan_mode_default settings (backend + UI) and the Slack `plan on/off/status` toggle. - enter_plan_mode returns a terminating ToolMessage, fixing the missing ToolMessage error that silently dropped plan mode mid-run. - PlanReview: defer Yjs provider/doc teardown so React StrictMode's dev remount doesn't destroy and then reuse the collaboration provider. - e2e plan_review spec asserts plan_mode actually engages. - LangSmith trace-url resolution is best-effort: bail before any API call when the tenant is unset, cache failures, log at debug. - Add `pnpm run dev:mock`: same-origin Vite HMR harness with a real LLM, Alice/Bob mock users, and a GitHub login picker. * docs(plan-mode): drop stale references to removed profile/team defaults The plan_mode middleware docstring and the approve/reject dispatch comment still described the profile/team plan_mode_default resolution that no longer exists; reword to match model-driven entry + the per-thread carry. * feat(plan-mode): let any reviewer edit the plan, not just comment Drop the owner/commenter split for the plan document: everyone with read access edits and comments alike (DefaultThreadStoreAuth "editor" for all, editor always editable until a decision, anyone seeds the empty doc). This matches the collab WS, which already relays frames to every readable user. Plan approval stays owner-gated. * test(plan-mode): assert plan-mode entry via the tool's success message plan_mode lives only in run state for tool gating; it is not a persisted thread-state channel, so the previous `values.plan_mode === true` poll could never pass. Assert instead that enter_plan_mode's success ToolMessage ("Plan mode is active …") lands in the thread — which only happens when the tool's Command applies cleanly, the exact regression this guards. --------- Co-authored-by: Johannes du Plessis <51395795+johannes117@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
2026-06-23 15:06:58 -04:00
class _FakeThreadsClient:
async def create(
self, *, thread_id: str, metadata: dict[str, Any], if_exists: str
) -> dict[str, Any]:
return {"thread_id": thread_id, "metadata": metadata}
async def update(self, *, thread_id: str, metadata: dict[str, Any]) -> dict[str, Any]:
return {"thread_id": thread_id, "metadata": metadata}
async def get(self, thread_id: str) -> dict[str, Any]:
return {"thread_id": thread_id, "metadata": {}}
class _FakeRunsClient:
def __init__(self) -> None:
self.configurable: dict[str, Any] | None = None
async def create(
self,
thread_id: str,
assistant_id: str,
*,
input: dict[str, Any],
config: dict[str, Any],
if_not_exists: str = "reject",
stream_mode: list[str] | None = None,
stream_resumable: bool = False,
) -> dict[str, str]:
self.configurable = config["configurable"]
return {"run_id": "run-id"}
class _FakeLangGraphClient:
def __init__(self) -> None:
self.threads = _FakeThreadsClient()
self.runs = _FakeRunsClient()
@pytest.fixture
def dashboard_run_client(monkeypatch: pytest.MonkeyPatch) -> _FakeLangGraphClient:
client = _FakeLangGraphClient()
async def fake_get_profile(login: str) -> dict[str, Any]:
return {}
async def fake_ensure_token(login: str) -> None:
return None
async def fake_resolve_email(login: str, profile: dict[str, Any]) -> str:
return "octo@example.com"
monkeypatch.setattr(thread_api, "langgraph_client", lambda: client)
monkeypatch.setattr(thread_api, "get_profile", fake_get_profile)
monkeypatch.setattr(thread_api, "_ensure_dashboard_github_token", fake_ensure_token)
monkeypatch.setattr(thread_api, "_resolve_run_email", fake_resolve_email)
return client
def _run_start_command(plan_mode: bool | None) -> dict[str, Any]:
configurable: dict[str, Any] = {}
if plan_mode is not None:
configurable["plan_mode"] = plan_mode
return {
"method": "run.start",
"params": {
"input": {"messages": [{"role": "user", "content": "do work"}]},
"config": {"configurable": configurable},
},
}
def test_run_start_passes_plan_mode_when_enabled(
dashboard_run_client: _FakeLangGraphClient,
) -> None:
enriched = asyncio.run(
thread_api._enrich_run_start_command(
"thread-id",
"octo",
_run_start_command(True),
metadata={"source": "dashboard", "github_login": "octo"},
creating=False,
)
)
configurable = enriched["params"]["config"]["configurable"]
assert configurable["plan_mode"] is True
def test_run_start_omits_plan_mode_when_disabled(
dashboard_run_client: _FakeLangGraphClient,
) -> None:
enriched = asyncio.run(
thread_api._enrich_run_start_command(
"thread-id",
"octo",
_run_start_command(None),
metadata={"source": "dashboard", "github_login": "octo"},
creating=False,
)
)
configurable = enriched["params"]["config"]["configurable"]
assert "plan_mode" not in configurable
def test_thread_summary_reports_plan_mode() -> None:
summary = thread_api._thread_summary(
{"thread_id": "t1", "metadata": {"source": "dashboard", "plan_mode": True}}
)
assert summary["planMode"] is True
summary_off = thread_api._thread_summary(
{"thread_id": "t2", "metadata": {"source": "dashboard"}}
)
assert summary_off["planMode"] is False
def test_plan_mode_guidance_section_always_present() -> None:
"""The guidance section telling the agent about enter_plan_mode should be in every prompt."""
prompt = construct_system_prompt(working_dir="/work", plan_mode=False)
assert "enter_plan_mode" in prompt
assert "Plan Mode" in prompt
def test_plan_mode_guidance_section_present_when_enabled() -> None:
prompt = construct_system_prompt(working_dir="/work", plan_mode=True)
assert "enter_plan_mode" in prompt
assert "Plan Mode (ACTIVE)" in prompt
refactor: adopt modular webhook architecture (#1621) + port fork customizations (#85) * Adopt upstream modular webhook skeleton (#1621) Apply the durable-interrupt-dispatch refactor: split the monolithic webapp.py into a thin routing layer plus per-source handlers in webhooks/{github,slack,linear}.py, and add completion.py, dispatch.py, and reconcile.py. Reconcile fork divergence by keeping the Bedrock/ Fireworks cross-provider fallback, the no-agent-attribution prompt policy, the dashboard-handoff re-export, and the Slack channel-info cache. ci_autofix is restored on the new dispatch model in a later commit. Refs: #80 * Port fork webhook security delta onto modular handlers Re-apply the fork's security customizations that #1621 did not carry: Linear webhook replay protection (freshness window on the signed webhookTimestamp), per-repo token-cache binding threaded through the thread token resolvers, the INTERNAL_BOT_LOGINS self-check in the review-finding-reply path, and a user-mapping cache refresh before email resolution on the issue and PR-comment paths (multi-replica staleness). Existing fork security tests pass unchanged. Refs: #80 * Restore CI auto-fix on the modular dispatch model Bring back ci_autofix.py and the ci_monitor graph that #1621 deleted, re-wiring the fork's security-reviewed PR-babysitting onto the new structure: the CI-event, autofix-toggle, and review-feedback handlers move into webhooks/github.py and the github_webhook router re-gains the check_run/check_suite/workflow_run/status routing plus the autofix command and actionable-review branches. Auto-fix runs now dispatch through dispatch_agent_run (durability + completion webhook) while keeping the deliberate batch-while-busy skip-rule via get_thread_active_status. Restore langgraph.json's ci_monitor entry and the fork autofix tests (dispatch mock + import paths re-pointed). Refs: #80 * Reformat and update docs for the modular webhook split Point CLAUDE.md and deploy/MIGRATION.md at the new webhooks/ modules and the dispatch/completion/reconcile contract, and mark the user-mapping cache-refresh fix as applied on the GitHub handlers. Refs: #80 * Restore reject backstop for autofix dispatch A burst of near-simultaneous CI events for one head SHA can slip past the busy-check before the dedupe SHA is recorded, so dispatch the autofix path with multitask_strategy=reject (dev's prior platform default) to drop duplicate concurrent creates instead of letting them interrupt each other. Also make the completion failure-reply dedup claim-then-post and drop the unreachable interrupted branch. --------- Co-authored-by: amoussa1229 <166072409+amoussa1229@users.noreply.github.com>
2026-06-30 18:46:46 -04:00
async def test_enter_plan_mode_tool_returns_command() -> None:
feat: plan mode with model-driven entry and collaborative review (#1580) * feat: add plan mode for read-only research and planning Adds a per-run plan_mode flag that puts the agent in a read-only research phase: a strong prompt section is injected and mutating tools are stripped via ExcludeToolsMiddleware so the agent proposes a reviewable implementation plan before any edits. Surfaced in the dashboard UI with a Plan toggle (Shift+Tab) wired through the thread API. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: enforce plan-mode read-only at tool layer and disable subagents Addresses PR review: plan mode previously relied on prompt text to keep the shell read-only and left the task subagent (built with its own write/PR/Linear tools) unrestricted. Now `task` is excluded so research cannot be delegated to a mutating subagent, and a new PlanModeShellGuardMiddleware enforces a read-only command allowlist on `execute`, blocking writes, git state changes, installs, redirection, and command substitution regardless of model/prompt-injection compliance. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: harden plan-mode shell guard against wrapped mutations Block git global options that take values (-C, --git-dir, ...) from being misread as the subcommand, reject config-injection options (-c, --config-env, --exec-path), and drop the env command wrapper that could run arbitrary commands. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * feat: add plan mode with enter_plan_mode tool, profile/team defaults, Slack commands and approval flow - enter_plan_mode tool: agent self-activates plan mode via Command(update={'plan_mode': True}) - Plan mode resolution: per-thread > profile default > team default > False - PLAN_MODE_GUIDANCE_SECTION: always-present prompt section telling agent about the tool - profile_plan_mode_default and team plan_mode_default settings - Slack plan on/off/status commands with thread metadata persistence - slack_thread_reply plan_approval=True renders Approve/Revise/Cancel buttons - Interactivity handler: approve triggers implementation run, cancel posts confirmation - Frontend: plan_mode_default in Profile/ProfileUpdate/TeamSettings types and UI toggles Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * test: add tests for enter_plan_mode tool, profile/team defaults, Slack plan commands, approval blocks Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * refactor(plan-mode): drop shell guard, rely on prompt for read-only discipline Remove PlanModeShellGuardMiddleware and its enforcement of read-only shell commands during plan mode. Plan mode now relies on the system prompt to instruct the agent not to run mutating commands; the mutating-tool exclusion (ExcludeToolsMiddleware) is retained. * test(open-swe): add Playwright E2E for the Slack → PR → web handoff Local, secrets-free end-to-end suite that drives the full happy path through mock Slack/GitHub control panels and the real dashboard UI. Only the LLM and external SaaS HTTP boundaries (GitHub/Slack APIs, OAuth token mint) are faked — the real process_slack_mention, get_agent, deepagents loop, tools, middleware, and dashboard authorization all run under `langgraph dev` with a scripted fake chat model and a local temp-dir sandbox. - full_flow: a Slack mention runs the agent, which implements a change in the sandbox, opens a PR against a fake GitHub remote, and replies with the PR link in the same thread. - dashboard: clicking the bot's real "Open in Web" link loads the built ui/ app (served same-origin); the thread owner can continue the conversation, while a different user sees the same thread read-only (no composer). Wired into Agent CI as a `Playwright E2E` job that runs on pull requests. * fix(open-swe): serve E2E UI assets via explicit route; pin Playwright The dashboard E2E served the built ui/ SPA's /assets via app.mount(StaticFiles), but LangGraph's custom-app loader serves APIRoutes and drops sub-app Mounts, so /assets 404'd under `langgraph dev` in CI — the React app never booted and the composer/transcript never rendered. Serve assets via an explicit route instead. Also pin @playwright/test to the latest (1.61.0) for reproducible runs, and make the owner composer assertion tolerant of either hydration state. * test(open-swe): record Playwright trace + video on every E2E run Capture a replayable trace (DOM snapshots, network, console, source) and a screen recording for every test, not just retries, plus a screenshot on failure. The CI job already uploads playwright-report/ and test-results/, so each run now has a downloadable replay; documented how to open it. * feat(plan-mode): collaborative plan review with BlockNote + Yjs When the agent enters plan mode it writes the plan as a markdown file in the sandbox (save_plan tool), publishes it, and posts a review link to the source channel. Reviewers open the plan inside the dashboard (under the /agents shell), read it rendered in a BlockNote editor, and leave inline comments synced live over Yjs. Only the thread owner can approve; any reviewer can request changes. On approve/reject the comments are harvested and handed to the agent for the follow-up run; the agent never sees comments mid-review. - agent: enter_plan_mode persists plan state; new save_plan tool; prompt shares the plan-review link. - dashboard: Yjs WebSocket collab server (pycrdt-websocket) with store-backed snapshots; plan content/status store; plan REST API (get/approve/reject, owner-only approve, client-harvested comments); planStatus on thread summaries. - ui: BlockNote native comments (CommentsExtension + YjsThreadStore) plan page mounted under the agents shell, with a "Review plan" banner in the thread view and a back-link; theme-aware (dark mode) using the dashboard tokens. - e2e: Playwright coverage of the full Slack -> plan -> review -> approve -> PR flow, including cross-user comment sync and owner-only approval. * fix(plan-mode): address review feedback (authz, overrides, leaks, deps) - plan-collab WS: authorize per-thread before joining a room (same read gate as the REST API) — previously any logged-in user could join any thread (IDOR). - plan-collab: tie the snapshot flusher to active connections (refcount) so each opened plan no longer leaks a permanent 1.5s task on the shared event loop. - plan decisions: include thread_id in the follow-up run configurable so the run resumes the existing thread; set plan_mode explicitly so approve forces it off. - get_agent: an explicit per-thread plan_mode (Slack `plan off`, approved plan, dashboard toggle) now overrides profile/team defaults instead of falling back. - plan mode tool gating moved to a state-aware PlanModeMiddleware installed unconditionally, so a mid-run enter_plan_mode restricts the next model turn; before_agent resets stale plan_mode so a later run isn't forced back into it. - exclude write-capable http_request from plan mode. - pin pycrdt / pycrdt-websocket with upper bounds. Includes the latest base (#1583): E2E UI assets served via explicit route (fixes the Playwright CI failure — LangGraph's app loader drops sub-app mounts). * style: ruff format plan_collab.py * fix(plan-mode): owner-gate Slack approval + same-origin check on collab WS - Slack "Approve & Implement" now verifies the clicking user is the plan requester (owner, via the stored triggering_user_id) before implementing — matching the dashboard API's owner-only approval. Non-owners are pointed to Revise / feedback. - The plan-collab WebSocket validates the handshake Origin against the dashboard allowlist before accept() (no-op when unconfigured, e.g. local/dev), mirroring the REST require_same_origin CSRF defense. * fix(plan-mode): enter plan mode only via the model + local mock dev harness Plan mode is now entered solely when the model calls enter_plan_mode. Removed the per-user and team plan_mode_default settings (backend + UI) and the Slack `plan on/off/status` toggle. - enter_plan_mode returns a terminating ToolMessage, fixing the missing ToolMessage error that silently dropped plan mode mid-run. - PlanReview: defer Yjs provider/doc teardown so React StrictMode's dev remount doesn't destroy and then reuse the collaboration provider. - e2e plan_review spec asserts plan_mode actually engages. - LangSmith trace-url resolution is best-effort: bail before any API call when the tenant is unset, cache failures, log at debug. - Add `pnpm run dev:mock`: same-origin Vite HMR harness with a real LLM, Alice/Bob mock users, and a GitHub login picker. * docs(plan-mode): drop stale references to removed profile/team defaults The plan_mode middleware docstring and the approve/reject dispatch comment still described the profile/team plan_mode_default resolution that no longer exists; reword to match model-driven entry + the per-thread carry. * feat(plan-mode): let any reviewer edit the plan, not just comment Drop the owner/commenter split for the plan document: everyone with read access edits and comments alike (DefaultThreadStoreAuth "editor" for all, editor always editable until a decision, anyone seeds the empty doc). This matches the collab WS, which already relays frames to every readable user. Plan approval stays owner-gated. * test(plan-mode): assert plan-mode entry via the tool's success message plan_mode lives only in run state for tool gating; it is not a persisted thread-state channel, so the previous `values.plan_mode === true` poll could never pass. Assert instead that enter_plan_mode's success ToolMessage ("Plan mode is active …") lands in the thread — which only happens when the tool's Command applies cleanly, the exact regression this guards. --------- Co-authored-by: Johannes du Plessis <51395795+johannes117@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
2026-06-23 15:06:58 -04:00
from langchain_core.messages import ToolMessage
from langchain_core.tools import tool as as_tool
from langgraph.types import Command
from agent.tools.enter_plan_mode import enter_plan_mode
# Wrap as the agent does so the InjectedToolCallId is supplied from the call.
wrapped = as_tool(enter_plan_mode)
refactor: adopt modular webhook architecture (#1621) + port fork customizations (#85) * Adopt upstream modular webhook skeleton (#1621) Apply the durable-interrupt-dispatch refactor: split the monolithic webapp.py into a thin routing layer plus per-source handlers in webhooks/{github,slack,linear}.py, and add completion.py, dispatch.py, and reconcile.py. Reconcile fork divergence by keeping the Bedrock/ Fireworks cross-provider fallback, the no-agent-attribution prompt policy, the dashboard-handoff re-export, and the Slack channel-info cache. ci_autofix is restored on the new dispatch model in a later commit. Refs: #80 * Port fork webhook security delta onto modular handlers Re-apply the fork's security customizations that #1621 did not carry: Linear webhook replay protection (freshness window on the signed webhookTimestamp), per-repo token-cache binding threaded through the thread token resolvers, the INTERNAL_BOT_LOGINS self-check in the review-finding-reply path, and a user-mapping cache refresh before email resolution on the issue and PR-comment paths (multi-replica staleness). Existing fork security tests pass unchanged. Refs: #80 * Restore CI auto-fix on the modular dispatch model Bring back ci_autofix.py and the ci_monitor graph that #1621 deleted, re-wiring the fork's security-reviewed PR-babysitting onto the new structure: the CI-event, autofix-toggle, and review-feedback handlers move into webhooks/github.py and the github_webhook router re-gains the check_run/check_suite/workflow_run/status routing plus the autofix command and actionable-review branches. Auto-fix runs now dispatch through dispatch_agent_run (durability + completion webhook) while keeping the deliberate batch-while-busy skip-rule via get_thread_active_status. Restore langgraph.json's ci_monitor entry and the fork autofix tests (dispatch mock + import paths re-pointed). Refs: #80 * Reformat and update docs for the modular webhook split Point CLAUDE.md and deploy/MIGRATION.md at the new webhooks/ modules and the dispatch/completion/reconcile contract, and mark the user-mapping cache-refresh fix as applied on the GitHub handlers. Refs: #80 * Restore reject backstop for autofix dispatch A burst of near-simultaneous CI events for one head SHA can slip past the busy-check before the dedupe SHA is recorded, so dispatch the autofix path with multitask_strategy=reject (dev's prior platform default) to drop duplicate concurrent creates instead of letting them interrupt each other. Also make the completion failure-reply dedup claim-then-post and drop the unreachable interrupted branch. --------- Co-authored-by: amoussa1229 <166072409+amoussa1229@users.noreply.github.com>
2026-06-30 18:46:46 -04:00
result = await wrapped.ainvoke(
feat: plan mode with model-driven entry and collaborative review (#1580) * feat: add plan mode for read-only research and planning Adds a per-run plan_mode flag that puts the agent in a read-only research phase: a strong prompt section is injected and mutating tools are stripped via ExcludeToolsMiddleware so the agent proposes a reviewable implementation plan before any edits. Surfaced in the dashboard UI with a Plan toggle (Shift+Tab) wired through the thread API. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: enforce plan-mode read-only at tool layer and disable subagents Addresses PR review: plan mode previously relied on prompt text to keep the shell read-only and left the task subagent (built with its own write/PR/Linear tools) unrestricted. Now `task` is excluded so research cannot be delegated to a mutating subagent, and a new PlanModeShellGuardMiddleware enforces a read-only command allowlist on `execute`, blocking writes, git state changes, installs, redirection, and command substitution regardless of model/prompt-injection compliance. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * fix: harden plan-mode shell guard against wrapped mutations Block git global options that take values (-C, --git-dir, ...) from being misread as the subcommand, reject config-injection options (-c, --config-env, --exec-path), and drop the env command wrapper that could run arbitrary commands. Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * feat: add plan mode with enter_plan_mode tool, profile/team defaults, Slack commands and approval flow - enter_plan_mode tool: agent self-activates plan mode via Command(update={'plan_mode': True}) - Plan mode resolution: per-thread > profile default > team default > False - PLAN_MODE_GUIDANCE_SECTION: always-present prompt section telling agent about the tool - profile_plan_mode_default and team plan_mode_default settings - Slack plan on/off/status commands with thread metadata persistence - slack_thread_reply plan_approval=True renders Approve/Revise/Cancel buttons - Interactivity handler: approve triggers implementation run, cancel posts confirmation - Frontend: plan_mode_default in Profile/ProfileUpdate/TeamSettings types and UI toggles Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * test: add tests for enter_plan_mode tool, profile/team defaults, Slack plan commands, approval blocks Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> * refactor(plan-mode): drop shell guard, rely on prompt for read-only discipline Remove PlanModeShellGuardMiddleware and its enforcement of read-only shell commands during plan mode. Plan mode now relies on the system prompt to instruct the agent not to run mutating commands; the mutating-tool exclusion (ExcludeToolsMiddleware) is retained. * test(open-swe): add Playwright E2E for the Slack → PR → web handoff Local, secrets-free end-to-end suite that drives the full happy path through mock Slack/GitHub control panels and the real dashboard UI. Only the LLM and external SaaS HTTP boundaries (GitHub/Slack APIs, OAuth token mint) are faked — the real process_slack_mention, get_agent, deepagents loop, tools, middleware, and dashboard authorization all run under `langgraph dev` with a scripted fake chat model and a local temp-dir sandbox. - full_flow: a Slack mention runs the agent, which implements a change in the sandbox, opens a PR against a fake GitHub remote, and replies with the PR link in the same thread. - dashboard: clicking the bot's real "Open in Web" link loads the built ui/ app (served same-origin); the thread owner can continue the conversation, while a different user sees the same thread read-only (no composer). Wired into Agent CI as a `Playwright E2E` job that runs on pull requests. * fix(open-swe): serve E2E UI assets via explicit route; pin Playwright The dashboard E2E served the built ui/ SPA's /assets via app.mount(StaticFiles), but LangGraph's custom-app loader serves APIRoutes and drops sub-app Mounts, so /assets 404'd under `langgraph dev` in CI — the React app never booted and the composer/transcript never rendered. Serve assets via an explicit route instead. Also pin @playwright/test to the latest (1.61.0) for reproducible runs, and make the owner composer assertion tolerant of either hydration state. * test(open-swe): record Playwright trace + video on every E2E run Capture a replayable trace (DOM snapshots, network, console, source) and a screen recording for every test, not just retries, plus a screenshot on failure. The CI job already uploads playwright-report/ and test-results/, so each run now has a downloadable replay; documented how to open it. * feat(plan-mode): collaborative plan review with BlockNote + Yjs When the agent enters plan mode it writes the plan as a markdown file in the sandbox (save_plan tool), publishes it, and posts a review link to the source channel. Reviewers open the plan inside the dashboard (under the /agents shell), read it rendered in a BlockNote editor, and leave inline comments synced live over Yjs. Only the thread owner can approve; any reviewer can request changes. On approve/reject the comments are harvested and handed to the agent for the follow-up run; the agent never sees comments mid-review. - agent: enter_plan_mode persists plan state; new save_plan tool; prompt shares the plan-review link. - dashboard: Yjs WebSocket collab server (pycrdt-websocket) with store-backed snapshots; plan content/status store; plan REST API (get/approve/reject, owner-only approve, client-harvested comments); planStatus on thread summaries. - ui: BlockNote native comments (CommentsExtension + YjsThreadStore) plan page mounted under the agents shell, with a "Review plan" banner in the thread view and a back-link; theme-aware (dark mode) using the dashboard tokens. - e2e: Playwright coverage of the full Slack -> plan -> review -> approve -> PR flow, including cross-user comment sync and owner-only approval. * fix(plan-mode): address review feedback (authz, overrides, leaks, deps) - plan-collab WS: authorize per-thread before joining a room (same read gate as the REST API) — previously any logged-in user could join any thread (IDOR). - plan-collab: tie the snapshot flusher to active connections (refcount) so each opened plan no longer leaks a permanent 1.5s task on the shared event loop. - plan decisions: include thread_id in the follow-up run configurable so the run resumes the existing thread; set plan_mode explicitly so approve forces it off. - get_agent: an explicit per-thread plan_mode (Slack `plan off`, approved plan, dashboard toggle) now overrides profile/team defaults instead of falling back. - plan mode tool gating moved to a state-aware PlanModeMiddleware installed unconditionally, so a mid-run enter_plan_mode restricts the next model turn; before_agent resets stale plan_mode so a later run isn't forced back into it. - exclude write-capable http_request from plan mode. - pin pycrdt / pycrdt-websocket with upper bounds. Includes the latest base (#1583): E2E UI assets served via explicit route (fixes the Playwright CI failure — LangGraph's app loader drops sub-app mounts). * style: ruff format plan_collab.py * fix(plan-mode): owner-gate Slack approval + same-origin check on collab WS - Slack "Approve & Implement" now verifies the clicking user is the plan requester (owner, via the stored triggering_user_id) before implementing — matching the dashboard API's owner-only approval. Non-owners are pointed to Revise / feedback. - The plan-collab WebSocket validates the handshake Origin against the dashboard allowlist before accept() (no-op when unconfigured, e.g. local/dev), mirroring the REST require_same_origin CSRF defense. * fix(plan-mode): enter plan mode only via the model + local mock dev harness Plan mode is now entered solely when the model calls enter_plan_mode. Removed the per-user and team plan_mode_default settings (backend + UI) and the Slack `plan on/off/status` toggle. - enter_plan_mode returns a terminating ToolMessage, fixing the missing ToolMessage error that silently dropped plan mode mid-run. - PlanReview: defer Yjs provider/doc teardown so React StrictMode's dev remount doesn't destroy and then reuse the collaboration provider. - e2e plan_review spec asserts plan_mode actually engages. - LangSmith trace-url resolution is best-effort: bail before any API call when the tenant is unset, cache failures, log at debug. - Add `pnpm run dev:mock`: same-origin Vite HMR harness with a real LLM, Alice/Bob mock users, and a GitHub login picker. * docs(plan-mode): drop stale references to removed profile/team defaults The plan_mode middleware docstring and the approve/reject dispatch comment still described the profile/team plan_mode_default resolution that no longer exists; reword to match model-driven entry + the per-thread carry. * feat(plan-mode): let any reviewer edit the plan, not just comment Drop the owner/commenter split for the plan document: everyone with read access edits and comments alike (DefaultThreadStoreAuth "editor" for all, editor always editable until a decision, anyone seeds the empty doc). This matches the collab WS, which already relays frames to every readable user. Plan approval stays owner-gated. * test(plan-mode): assert plan-mode entry via the tool's success message plan_mode lives only in run state for tool gating; it is not a persisted thread-state channel, so the previous `values.plan_mode === true` poll could never pass. Assert instead that enter_plan_mode's success ToolMessage ("Plan mode is active …") lands in the thread — which only happens when the tool's Command applies cleanly, the exact regression this guards. --------- Co-authored-by: Johannes du Plessis <51395795+johannes117@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
2026-06-23 15:06:58 -04:00
{"name": "enter_plan_mode", "args": {}, "id": "call-1", "type": "tool_call"}
)
assert isinstance(result, Command)
assert result.update["plan_mode"] is True
messages = result.update["messages"]
assert len(messages) == 1
assert isinstance(messages[0], ToolMessage)
assert messages[0].tool_call_id == "call-1"
def test_enter_plan_mode_exported() -> None:
from agent.tools import enter_plan_mode
assert callable(enter_plan_mode)
def test_build_plan_approval_blocks_has_three_buttons() -> None:
from agent.tools.slack_thread_reply import _build_plan_approval_blocks
blocks = _build_plan_approval_blocks("Here is my plan")
assert len(blocks) == 2
assert blocks[0]["type"] == "section"
actions = blocks[1]
assert actions["type"] == "actions"
elements = actions["elements"]
assert len(elements) == 3
texts = [e["text"]["text"] for e in elements]
assert "Approve & Implement" in texts
assert "Revise Plan" in texts
assert "Cancel" in texts
def test_build_plan_approval_blocks_values_have_plan_approval_type() -> None:
import json
from agent.tools.slack_thread_reply import _build_plan_approval_blocks
blocks = _build_plan_approval_blocks("plan text")
for element in blocks[1]["elements"]:
value = json.loads(element["value"])
assert value["type"] == "plan_approval"
assert value["action"] in ("approve", "revise", "cancel")