Skip to content

unxed/f4 / .agents/skills/aif-verify

aif-verify

Verified

Verify completed implementation against the plan. Checks that all tasks were implemented, nothing was forgotten, code compiles, tests pass, and quality standards are met. Use after "/aif-implement" completes, or when user says "verify", "check work", "did we miss anything".

FileKindSize
SKILL.mdskill md25.0 KB
references/CONTEXT-GATES-AND-OWNERSHIP.mdmarkdown10.7 KB
references/GATE-RESULT-CONTRACT.mdmarkdown3.7 KB

references/GATE-RESULT-CONTRACT.md

View on GitHub
# Machine-Readable Quality Gate Result Contract

Quality gate skills keep their human-readable markdown report first, then append one final fenced JSON block for orchestrators.

Orchestrators must parse the last `aif-gate-result` fenced block in the final gate output. Earlier fences in documentation, examples, or quoted prior output are illustrative and are not the gate result.

Supported gates:
- `verify` for `/aif-verify`
- `review` for `/aif-review`
- `security` for `/aif-security-checklist`
- `rules` for `/aif-rules-check`

The fenced block language must be exactly `aif-gate-result`.

```aif-gate-result
{
  "schema_version": 1,
  "gate": "verify",
  "status": "fail",
  "blocking": true,
  "blockers": [
    {
      "id": "verify-task-1",
      "severity": "error",
      "file": "src/example.ts",
      "summary": "Required behavior is missing."
    }
  ],
  "affected_files": ["src/example.ts"],
  "suggested_next": {
    "command": "/aif-fix",
    "reason": "Blocking implementation gaps remain."
  }
}
```

## Schema

```text
"schema_version": 1
"gate": "verify|review|security|rules"
"status": "pass|warn|fail"
"blocking": true|false
"blockers": [
  {
    "id": "stable-finding-id",
    "severity": "error|warning",
    "file": "optional/path",
    "summary": "short human-readable summary"
  }
]
"affected_files": ["path/to/file"]
"suggested_next": {
  "command": "/aif-fix|/aif-rules|/aif-architecture|/aif-roadmap|/aif-commit|null",
  "reason": "short rationale or null"
}
```

Rules:
- The fenced block contains JSON only. Do not include comments, trailing commas, Markdown bullets, or prose inside it.
- `status` must be one of `pass`, `warn`, or `fail`.
- `blocking` must be a boolean.
- `blockers` contains only findings that should block the current quality gate. Non-blocking notes stay in the human summary.
- `blockers[].severity` uses the shared gate scale: `error` for blocking findings and `warning` only when policy treats a warning-class finding as blocking. Security domain severities map to this scale (`critical`/`high` -> `error`; `medium`/`low` -> non-blocking human warning unless policy escalates them).
- `affected_files` is a predictable top-level array of files the gate actually evaluated or cited in the result. It is not limited to blocker files, and it should not include unrelated repository files. Use an empty array when no specific files apply.
- `suggested_next.command` must come from the global allowlist: `/aif-fix`, `/aif-rules`, `/aif-architecture`, `/aif-roadmap`, `/aif-commit`, or `null`. Each gate may document a narrower subset for its own outputs.
- The JSON block appears after the human-readable summary so manual workflows remain readable.

## Status Semantics

- `pass`: the gate completed without blocking or warning findings.
- `warn`: the gate completed with non-blocking findings, missing optional context, ambiguous evidence, accepted skips, or advisory security/rules notes.
- `fail`: the gate found at least one blocker that should stop commit, merge, deploy, or implementation handoff.

## Suggested Next Command

Allowed commands by gate:
- `verify`: `/aif-fix`, `/aif-rules`, `/aif-architecture`, `/aif-roadmap`, `/aif-commit`, or `null`.
- `review`: `/aif-fix`, `/aif-rules`, `/aif-architecture`, `/aif-roadmap`, `/aif-commit`, or `null`.
- `security`: `/aif-fix` or `null`.
- `rules`: `/aif-rules`, `/aif-fix`, or `null`.

- `/aif-fix`: code, tests, config, or documentation must be corrected.
- `/aif-rules`: rules are missing, ambiguous, or need a writer-owned update.
- `/aif-architecture`: architecture context is stale or violated.
- `/aif-roadmap`: roadmap context is stale, missing, or contradicted.
- `/aif-commit`: the gate is clean and committing is the natural next step.
- `null`: no AI Factory command is appropriate.
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall