Scope note: This score reflects structural compliance with the spec format as assessed by an AI model against a structured scoring rubric intended to be applied consistently. It does not guarantee real-world session effectiveness or consistent LLM behavior across providers. Treat it as a structural checklist result, not a quality certification.
File: AGENTS.md
Validated: 2026-05-08
Validator: skills/validate-agents-md
- Status: PASS
- Compliance Score: 100/100
- Score: 20/20
- Status: PASS
- Target file exists and is structurally complete.
<!-- spec_version: 1.4.2 -->comment is present and current — no version warning.- Purpose, precedence model, and repository map are explicitly defined.
- Heading structure is scannable and consistent.
- Score: 35/35
- Status: PASS
- Session-state summary is explicit.
- No-silent-transitions rule is present.
- Ambiguity rule is present.
- Command namespace policy and
/ack.*command reference are complete. - Default state values are explicitly defined.
- Formatting/path stability rules are present.
- Drift-control rule is present.
- Nested
AGENTS.mdprecedence policy is present.
- Score: 19/20
- Status: PASS
- Guidance is actionable and testable.
- No conflicting policy statements detected.
- Spec duplication is controlled.
- Minor opportunity: tighten a few descriptive lines in
Repository Project Contextto keep the operational profile lean.
- Score: 15/15
- Status: PASS
- Language remains provider-neutral.
- Guidance is runtime/tool portable.
- No mandatory single-runtime assumptions found.
- Score: 10/10
- Status: PASS
- Relative links in
AGENTS.mdresolve. - Referenced paths exist.
- Canonical path guidance aligns with repository reality (AGENTS-first posture).
- Clear AGENTS-first authority model with explicit precedence.
- Strong operational contract (state, transitions, command namespace).
- High repository alignment and link integrity.
- None.
- None.
- None.
- No blocking changes required.
- Keep validation re-runs coupled with future AGENTS structural edits.
- If section growth continues, split non-operational detail into referenced docs while keeping AGENTS policy-focused.
- No recurring failures detected in this run.