ade/missions removal 0a31b5b5#359
Conversation
arul28
commented
May 25, 2026
- Remove missions feature
- Clean up stale mission/orchestrator references in agent tooling and action schemas
- ship: prepare lane for review
|
Too many files changed for review. ( |
|
The latest updates on your projects. Learn more about Vercel for GitHub. |
|
Important Review skippedToo many files! This PR contains 298 files, which is 148 over the limit of 150. To get a review, narrow the scope: ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (2)
📒 Files selected for processing (298)
You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
@copilot review but do not make fixes |
|
Capy auto-review is paused for this organization because the monthly auto-review limit has been reached. Increase the limit or turn it off in billing settings to resume automatic reviews. |
…ction schemas Remove dead orchestrator action schemas from the automation editor (handlers no longer exist), and purge leftover mission references from .claude/commands and agent skill definitions that would mislead automated tooling. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
1a55f8a to
a95d9cc
Compare
|
@copilot review but do not make fixes |
There was a problem hiding this comment.
PR Review
Scope: 447 file(s), +1488 / −145659
Verdict: Needs changes
This PR removes the Missions product surface end-to-end (desktop services, orchestrator, CLI/RPC tools, iOS schema/docs, marketing assets) and folds automation defaults toward agent-session / built-in execution. The deletion scope is consistent, but legacy on-disk automation config is not fully migrated, which can block project config for existing users.
🐛 Functionality
[High] Legacy mission automations can make project config invalid
File: apps/desktop/src/main/services/config/projectConfigService.ts:L534-L540,L2226-L2228,L2891-L2896
Issue: coerceAutomationExecution no longer accepts kind: "mission", so stored rules with that execution kind lose their execution block on load. resolveEffectiveConfig then substitutes { kind: "agent-session", session: {} }, but many mission rules only set execution.mission.title and never set prompt. Validation then requires a prompt for every non-built-in automation, so validation.ok becomes false.
Repro: In .ade/ade.yaml, keep a rule like:
automations:
- id: legacy-mission
trigger: { type: manual }
execution:
kind: mission
mission: { title: "Run nightly" }Open the project (or call getEffective() / getExecutableConfig()). Config validation fails with agent-session automations require a prompt. and executable config load throws ADE_CONFIG_INVALID, which blocks automations and other paths that require executable config.
Fix: Add an explicit migration when kind is mission (or when launch-mission was the only action): map to agent-session, copy execution.mission.title / former action title into prompt (and execution.session.title if desired), and only then run validation. Alternatively, treat legacy mission rules as built-in until a prompt is authored.
[Medium] Legacy MissionModelConfig YAML is silently dropped
File: apps/desktop/src/main/services/config/projectConfigService.ts:L466-L474,L764-L765
Issue: Automation rules now use flat ModelConfig, but coerceModelConfig only reads top-level modelId. Existing configs shaped as { orchestratorModel: { modelId, thinkingLevel }, profileId } coerce to undefined, so automations lose their configured model with no validation error and fall back to defaults.
Repro: Load a project whose automation still has:
modelConfig:
orchestratorModel:
modelId: anthropic/claude-sonnet-4-6
thinkingLevel: mediumInspect effective config: modelConfig is missing for that rule.
Fix: In coerceModelConfig (or a one-time normalizer), accept legacy orchestratorModel and map it to flat ModelConfig (optionally log when profileId is ignored).
[Medium] launch-mission actions are stripped from built-in chains
File: apps/desktop/src/main/services/config/projectConfigService.ts:L87-L93,L477-L483
Issue: launch-mission was removed from AUTOMATION_ACTION_TYPE_SET, so legacy built-in rules whose only step was launch-mission load with an empty builtIn.actions list after coercion. That triggers Built-in automations need at least one task. validation errors, or—if execution also collapses to agent-session—the missing-prompt failure above.
Repro: Keep a rule with execution.kind: built-in and actions: [{ type: launch-mission, sessionTitle: "Nightly" }] only; reload config and observe validation failure or an empty action chain.
Fix: During coercion, rewrite launch-mission to agent-session execution (with prompt/title migration) instead of dropping the action.
Notes
- Runtime normalization in
automationService.normalizeRuntimeRulealready maps non-built-inexecution toagent-session, so the main gap is config load/validation, not dispatch of already-normalized rules. - Removing
/missionsroutes is handled safely: unknown paths fall through to/work(App.tsxcatch-all), so old bookmarks degrade instead of crashing. - Orphan
missions*/orchestrator_*tables are left in existing SQLite DBs (create paths removed, no drop migration). That is acceptable for this removal pass but worth a follow-up cleanup migration if you want smaller DBs and fewer dead FK targets.
Sent by Cursor Automation: BUGBOT in Versic
Reviewed only, no fixes applied. Audit scope: the last 3 commits ( Validation run:
What I checked:
Findings: no new blocking issues found in the reviewed changes; current state looks consistent with the removal and cleanup intent at |
Reviewed only; no fixes applied. I re-audited the current last 4 commits ( Checks run:
CI log review (run No blocking issues found in the reviewed code paths. |
|
@copilot review but do not make fixes |
Reviewed only; no fixes applied. I re-audited the current last 5 commits ( Checks run:
CI review: run No blocking issues found in the reviewed code paths. |

