Researches real projects, papers, and documented failures before proposing a new feature, architecture, or technology choice. Use for design decisions with multiple valid approaches, not for implementing an existing spec or fixing a known bug.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
Other people's documented failures count as observed failure. Before proposing a design, find out how others solved the same problem and where it broke for them.
vibe-decision-journal first)Frame the question — What exactly is being decided? What are the hard constraints (scale, latency, cost, team skills, existing stack)?
Decompose into sub-problems that can be researched independently.
Research each sub-problem using whatever web search or research tool the harness provides. Run sub-problems in parallel (subagents) when the harness supports it. For each, find:
| Requirement | Minimum |
|---|---|
| Real projects or production systems | 2+, with a link |
| Paper, post-mortem, or engineering write-up | 1+, with a link and year |
| Mechanism | How it actually works, not marketing copy |
| Known failures | What went wrong for others, and under what conditions |
| Fit | How it maps to our constraints, specifically |
Verify before citing — Every source must be something you actually opened in this session. Don't quote adoption numbers (stars, downloads, users) from memory; either read them from the source or leave them out. If search is unavailable, say so and label the output "from prior knowledge, unverified."
Synthesize — Compare the options side by side, with the imported failures up front.
Recommend — Base the recommendation on the evidence and state your confidence. Name the condition that would change your mind.
Record — Log the decision and the rejected alternatives with vibe-decision-journal.
| Thought | Reality |
|---|---|
| "I already know the best approach" | You know one approach. Research finds the others. |
| "This is too niche for research" | If someone built it, someone documented where it failed. |
| "Research will slow us down" | A rewrite is slower. |
| "Everyone uses X" | Popularity is not fitness for your constraints. |
| "I'll cite what I remember" | Unverified citations are the most common way research output goes wrong. |
| Criteria | Option A | Option B | Option C |
|---|---|---|---|
| Mechanism | |||
| Known failures | |||
| Fit with constraints |
[Evidence-based recommendation, confidence level, and what would change it]
[Links actually opened, one line each on what they showed]
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Validates completed work against defined acceptance criteria. Use after completing a task that has specific success criteria defined in issues, specs, or task descriptions.
The English summary is being prepared. Showing the original description.
Generates edge case, failure mode, and spec-driven test cases. Covers boundary values, nil inputs, concurrency, resource exhaustion, malformed data, and requirement-linked traceability tests. Use after happy-path tests exist and before claiming coverage is complete, when requirements lack tests, or before a security review.
The English summary is being prepared. Showing the original description.
Catches shortcuts and reward hacking — weakening or deleting tests to make them pass, hard-coding expected outputs, silently dropping requirements, or claiming work is done without running it. Use before declaring a task complete and whenever a test or requirement feels like it's in the way.
The English summary is being prepared. Showing the original description.
Safely integrates commits from parallel agent branches using sequential cherry-pick. Use after parallel work completes in isolated branches or worktrees.
The English summary is being prepared. Showing the original description.
Audits tests for concurrency safety — race conditions, shared mock state, cleanup ordering. Use when writing tests that involve goroutines, async operations, or shared mutable state.
The English summary is being prepared. Showing the original description.
Enforces tiered test coverage standards with three dimensions — line coverage by tier, spec-to-test traceability, and spec-to-code implementation mapping. Use before claiming code is complete.
The English summary is being prepared. Showing the original description.