ticket-review
Skill1 starsChanged 4 months agoMITView on GitHub
Use after triage to produce task files - analyzes requirements from a plain description or a Jira, Azure DevOps, or GitHub Issues ticket, and generates execution plans.
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 3 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Create this TODO verbatim and mark each item done as it completes:
- Distinctive: line 32
Use the triage classification.
- Distinctive: line 43
Keep exploration focused on the affected area.
- Very specific: line 72
Plain-text SIMPLE: write one task inline using the SIMPLE template in skills/task-generation/templates/task-files.md: TASK, DONE WHEN, …
- Very specific: line 77
Keep one task per logical change, under docs/tasks/{branch-slug}/ using docs/PROJECT_CONTEXT.md naming conventions.
- Very specific: line 82
Use skills/shared/dispatch.md for worker invocations at the matching process step:
- Very specific: line 86
Generation: after review approval and completeness, dispatch task-generation with the approved review path, exploration-log path, branch …
- Very specific: line 94
After task approval, ensure docs/tasks/{branch-slug}/exploration.md exists.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Ticket Review
Turn a triaged requirement into task files and an execution plan. Obtain separate user approvals for the requirement review and the generated tasks.
Process
Create this TODO verbatim and mark each item done as it completes:
TODO:
- [ ] Gather input (requirement source + context)
- [ ] Obtain requirement analysis - STANDARD/EPIC: worker result; SIMPLE: inline
- [ ] Review document ready (confirm/finalize) + user review gate
- [ ] Completeness gate
- [ ] Prepare task files - SIMPLE: inline; STANDARD/EPIC: worker result
- [ ] Obtain task-file approval and hand off the execution planStep 1: Gather Input
Obtain the plain requirement or ticket URL/key and any additional decisions, constraints, or priority notes. Capture the ticket reference at intake; its live content belongs to the analysis result.
Use the triage classification. Ticket references start at STANDARD; work touching auth, billing, migrations, tenant isolation, infra, or shared contracts requires at least STANDARD.
Derive the branch slug from the current branch by replacing / with -. Use the ticket ID as the search key when
available, otherwise the branch slug.
Step 2: Analyze the requirement
- STANDARD / EPIC: obtain the review-document path, findings, open markers, concerns, and any incremental result through the analysis handoff below.
- Plain-text SIMPLE: extract acceptance criteria and affected files/entry points inline, citing
file:line. Keep exploration focused on the affected area; this inline analysis serves as the review.
Present returned diff summaries. For up-to-date: yes, ask whether the user wants a full re-review. Apply a
recommended classification upgrade before generation. Relay blockers and specific clarification questions to the user.
Step 3: Review Document & User Review Gate
For STANDARD/EPIC, confirm the returned document against templates/review-document.md, including Last Reviewed
in UTC. Resolve gaps that would misdirect generation: missing sections, contradictions, material ambiguities, or scope
requiring decomposition. Judge readiness by decision completeness rather than prose polish.
Present the document, or SIMPLE inline analysis, for explicit approval of questions, assumptions, approach, scope, and security, compliance, or architectural concerns. Incorporate feedback through re-analysis and present it again.
Step 4: Completeness Gate
Proceed to generation only when every condition holds:
- User has approved the review document (Step 3)
- Every
[NEEDS CLARIFICATION]marker is resolved - Each acceptance criterion is verifiable by diff or command with a measurable expected outcome
- Each requirement covers edge cases, including empty input, missing data, and concurrent access
- Scope boundaries and exclusions are explicit
- Security implications are documented for auth/billing/tenant changes
Resolve unmet conditions with the user before continuing.
Step 5: Generate Task Files
- Plain-text SIMPLE: write one task inline using the SIMPLE template in
skills/task-generation/templates/task-files.md: TASK, DONE WHEN, VERIFY. - STANDARD / EPIC: obtain task paths, execution plan, coverage summary, and applicable QA/memory paths through the generation handoff below.
Keep one task per logical change, under docs/tasks/{branch-slug}/ using docs/PROJECT_CONTEXT.md naming conventions.
Limit task-file writes to the current branch's directory.
Handoff
Use skills/shared/dispatch.md for worker invocations at the matching process step:
- Analysis: dispatch
requirement-analysisfor STANDARD/EPIC with the requirement source, additional context, branch slug, and search key. Route review feedback back to it; revise SIMPLE analysis inline. - Generation: after review approval and completeness, dispatch
task-generationwith the approved review path, exploration-log path, branch slug, classification, and QA mode fromdocs/PROJECT_CONTEXT.md(off when absent). - Report: the task files + execution plan.
- Task approval: present the task files and plan in chat and wait for explicit user approval. Incorporate requested scope or ordering changes before implementation.
Dispatch execution
After task approval, ensure docs/tasks/{branch-slug}/exploration.md exists; create a missing log using
templates/exploration.md. Generate the driving TODO from the approved plan:
TODO:
- [ ] Implement task 001: <name> - dispatch execute-task subagent
- [ ] Verify task 001 - dispatch verify-task subagent
- [ ] Implement task 002: <name> - dispatch execute-task subagent
- [ ] Verify task 002 - dispatch verify-task subagent
...
- [ ] Run pr-review for full ticket - dispatch pr-review subagent
- [ ] If pr-review approves: invoke wrap-up- TODO update: execute the driving TODO through the shared dispatch contract, preserving the approved groups, independent workers, outcome handling, and bounded fix loop. Continue to wrap-up after pr-review approves.
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- elmoritz/agents-skills-setup★ 4same name
- SUDARSHANCHAUDHARI/QACraft★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository21 other files it gives its agents.
- address-feedbackSkill
- brainstormSkill
- code-conventionsSkill
- comprehensionSkill
- debugSkill
- definition-of-doneSkill
- execute-taskSkill
- migrate-docsSkill
- pr-reviewSkill
- repo-scanSkill
- requirement-analysisSkill
- securitySkill
- setupSkill
- start-ritusSkill
- syncSkill
- task-generationSkill
- tddSkill
- testing-policySkill
- triageSkill
- verify-taskSkill
- wrap-upSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

