quality-assurance
Skill98 starsChanged 26 days agoMITView on GitHub
Reference material with consistency-analysis heuristics and checklist-management rules. Loaded on demand by `analyze-compliance` and `quality-control`; not directly invokable.
11 other projects use a version of this.
Adopted by
Most recently taken up by
- thistleknot/skillstheir own version, first seen Nov 8, 2026
- PhoenixNest/AGENT-GLOBAL-BASEtheir own version, first seen Oct 20, 2026
- axross/skillstheir own version, first seen Oct 13, 2026
- jasonkneen/kirotheir own version, first seen Oct 8, 2026
- Ghosteken/agent-harnesstheir own version, first seen Oct 1, 2026
- brylie/compendiumtheir own version, first seen Sep 13, 2026
Other versions11 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 11
- jasonkneen/kiro★ 753same name
- liangdabiao/claude-data-analysis-ultra-main★ 282same name
- nahisaho/MUSUBI★ 74same name
- liangdabiao/claudesdk-data-chat★ 4same name
- Ghosteken/agent-harness★ 2same name
- PhoenixNest/AGENT-GLOBAL-BASE★ 2same name
- axross/skills★ 1same name
- brylie/compendium★ 1same name
- Luis85/agentic-workflow★ 0same name
- matheusflorindo32/apex-elite-framework★ 0same name
1 more not listed
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 repository73 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- sddp-amendSkill
- sddp-analyzeSkill
- sddp-autopilotSkill
- sddp-checklistSkill
- sddp-clarifySkill
- sddp-devopsSkill
- sddp-devsetupSkill
- sddp-implement-qc-loopSkill
- sddp-implementSkill
- sddp-initSkill
- sddp-planSkill
- sddp-prdSkill
- sddp-projectplanSkill
- sddp-qcSkill
- sddp-regenSkill
- sddp-reverseengineerSkill
- sddp-specifySkill
- sddp-systemdesignSkill
- sddp-tasksSkill
- sddp-amendSkill
- sddp-analyzeSkill
- sddp-autopilotSkill
- sddp-checklistSkill
- sddp-clarifySkill
- sddp-devopsSkill
- sddp-devsetupSkill
- sddp-implement-qc-loopSkill
- sddp-implementSkill
- sddp-initSkill
- sddp-planSkill
- sddp-prdSkill
- sddp-projectplanSkill
- sddp-qcSkill
- sddp-regenSkill
- sddp-reverseengineerSkill
- sddp-specifySkill
- sddp-systemdesignSkill
- sddp-tasksSkill
- adr-authoringSkill
- artifact-conventionsSkill
- clarification-strategiesSkill
- compact-communicationSkill
- evidence-led-authoringSkill
- implementation-standardsSkill
- instructions-managementSkill
- markdown-compressionSkill
- plan-authoringSkill
- sdd-methodologySkill
- sddp-amendSkill
- sddp-analyzeSkill
- sddp-autopilotSkill
- sddp-checklistSkill
- sddp-clarifySkill
- sddp-devopsSkill
- sddp-devsetupSkill
- sddp-implement-qc-loopSkill
- sddp-implementSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 6 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
When performing consistency analysis, verify the following relationships:
- Very specific: line 14
Entity Matching: Do entities defined in spec.md (Data Requirements) match the data-model.md?
- Distinctive: line 15
Complexity Check: If the spec marked a feature as "high complexity", does the plan include specific architectural mitigations?
- Distinctive: line 19
Phase Ordering: Do tasks respect the allowed phase order (optional Setup → optional Foundational → Story Phases → optional Polish)?
- Distinctive: line 42
Every generated checklist should consider these categories if relevant:
- Distinctive: line 49
Use the template at assets/checklist-template.md.
- Very specific: line 53
After a checklist is generated, the TestEvaluator sub-agent automatically evaluates every unchecked item against the feature artifacts.
- Distinctive: line 70
All P1 user stories or objectives have tasks.
- Very specific: line 72
If checklists exist in specs/<feature-folder>/checklists/, all must pass (or be explicitly overridden).
“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Quality Assurance Guide
Analysis Heuristics (/sddp-analyze)
When performing consistency analysis, verify the following relationships:
1. Spec vs. Plan Alignment
- Requirement Coverage: Does every P1/P2/P3 user story or objective in
spec.mdhave a corresponding section inplan.md? - Entity Matching: Do entities defined in
spec.md(Data Requirements) match thedata-model.md? - Complexity Check: If the spec marked a feature as "high complexity", does the plan include specific architectural mitigations?
2. Plan vs. Tasks Alignment
- Task Completeness: Does every section of the implementation plan have at least one corresponding task in
tasks.md? - Phase Ordering: Do tasks respect the allowed phase order (optional Setup → optional Foundational → Story Phases → optional Polish)?
- Missing Chunks: Are there major components in the architecture diagram that are absent from the task list?
3. Instructions Compliance
- Critical Violation: Any plan decision that contradicts
project-instructions.mdis a CRITICAL error. - Reporting: Flag these immediately with high severity.
Checklist Management (/sddp-checklist)
Checklists are the primary gate for the implementation phase.
Checklist Queue (.checklists)
The /sddp-plan agent generates a .checklists queue file in FEATURE_DIR/checklists/ that recommends checklist domains based on risk signals detected in the technical plan. This enables automated or sequential checklist generation without manual domain input.
- File:
FEATURE_DIR/checklists/.checklists - Format:
- [ ] CHL### Domain Name(e.g.,- [ ] CHL001 Security) - Generation:
/sddp-plancreates the file after the Post-Design Gate, capped by theMaxChecklistCountsetting in.github/sddp-config.md(default: 1). The plan agent may generate fewer entries if only fewer domains are relevant. - Consumption:
/sddp-checklistpicks the first unchecked entry when no explicit domain is provided, and marks it- [X]after completion. - Advisory, not blocking: The queue itself does not block implementation. Only the generated checklist files (
.md) inchecklists/are gating artifacts. - Override: An explicit domain argument to
/sddp-checklistalways takes priority over the queue.
Standard Categories
Every generated checklist should consider these categories if relevant:
- Security: Authz rules, input validation, secret handling.
- Performance: Indexing strategy, N+1 query checks, caching.
- Observability: Logging points, metric emission, error context.
- Testing: Unit test coverage, edge case handling.
Template
Use the template at assets/checklist-template.md.
Evaluation (TestEvaluator)
After a checklist is generated, the TestEvaluator sub-agent automatically evaluates every unchecked item against the feature artifacts. Each item receives one of three outcomes:
- PASS — The question is clearly answered by existing artifacts. Item is marked
[X]with an inline annotation citing the evidence source. - RESOLVE — The question reveals a genuine gap. The evaluator amends the relevant artifact(s) (e.g., adds missing
FR-###,TR-###,OR-###, orRR-###tospec.md, adds task totasks.md) then marks the item[X]. - ASK — The question is ambiguous or has multiple valid resolutions. The evaluator batches these, asks the user, then applies the chosen resolution.
Automated agents may change checkbox state from - [ ] to - [X] when supported by verified evidence or an explicit applied resolution.
It is invoked in two places:
/sddp-checklist: Automatically after checklist generation (Step 5)./sddp-implement: As a third gate option ("Auto-evaluate checklists now") when checklists fail the gate.
Definition of Done
Implementation Ready
A feature is "Implementation Ready" only when:
- Scale/Complexity risks are mitigated in
plan.md. - All P1 user stories or objectives have tasks.
- No "NEEDS CLARIFICATION" markers remain in Spec or Plan.
- If checklists exist in
specs/<feature-folder>/checklists/, all must pass (or be explicitly overridden).
Release Ready
A feature is "Release Ready" (eligible for .qc-passed) only when ALL of the following are true:
QC strictness: reads
## QC Strictnessfrom.github/sddp-config.md. Falls back to keyword scanning ofproject-instructions.md.
- All tests pass — unit and integration test suites report 0 failures.
- Coverage meets threshold — if
project-instructions.mddefines a coverage mandate, the measured coverage percentage meets or exceeds it. If no threshold is defined, coverage is reported but not enforced. - No CRITICAL or ERROR static analysis findings — linting and compilation issues at error severity are resolved.
- No CRITICAL security vulnerabilities — security scan findings classified as CRITICAL are resolved.
- All P1 work items PASSED — every P1 story or objective has its acceptance, validation, or verification criteria verified in the implementation.
- All Success Criteria (SC-###) PASSED — every success criterion is achievable by the current implementation.
- PI compliance: no violations — no
project-instructions.mdprinciples are violated. - No unresolved CRITICAL/ERROR
[BUG]tasks — CRITICAL/ERROR bugs may never be excluded with[DEFERRED]; only deferred WARNING bugs may remain unchecked. - No unacknowledged SKIPPED checks for PI-mandated categories — if
project-instructions.mdmandates any QC category (linting, security, coverage, accessibility, performance), that check must either pass, be explicitly acknowledged by the user as a risk-accepted waiver (WARNING), or generate FAIL with BUG tasks. Silent skips for PI-mandated categories do not satisfy Release Ready. - Manual verification attested — every required manual scenario records PASS with human identity, UTC timestamp, and evidence; pending or malformed attestation is BLOCKED.
- Current QC evidence —
.qc-passedreport and evidence SHA-256 digests match the currentqc-report.mdand every manifest file.
Note on criteria 2–4: These criteria apply when the respective tool ran. If a tool was SKIPPED and the category is PI-mandated, criterion 9 governs the outcome. If the category is not PI-mandated, a SKIPPED check does not block Release Ready but is reported as a WARNING recommendation.
This definition is enforced by /sddp-qc (Step 7) when determining the Overall Verdict.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
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.

