AI ecosystem audit rubric

Use this rubric only for the reusable, read-only audit of repository-controlled OpenCode and Claude Code assets.

Inputs

/ecosystem-audit [--scope=all|parity|skills|agents|commands|config] [--strict] [--relevance=gc-platform|portable]

Defaults: --scope=all, --relevance=gc-platform.

• all audits every dimension.
• parity audits mirrored OpenCode/Claude artifacts and necessary harness translations.
• skills, agents, commands, and config focus detailed analysis on that artifact class, retaining only reference checks necessary for interpretation.
• --strict promotes unverified references, incomplete mirrors, undocumented differences, and absent purpose/trigger/owner/output/mutation boundaries from risks/unknowns to findings. It never converts an inference into a confirmed fact.
• --relevance=gc-platform uses current repository stack markers and documented conventions. portable evaluates reusable harness design without requiring gc.platform-specific fit.

Audit boundary

Audit repository-controlled assets under .opencode/ and .claude/, plus project stack markers necessary to establish relevance. Exclude user-home configuration, secrets and secret-bearing files, node_modules, generated/build/cache directories, runtime state/runtime configuration, dependency lockfile contents, and product-source quality.

The audit is read-only: no edits, generated files, installs, hook execution, commits, remote actions, or delegation. Redact sensitive values.

Evidence standard

1. Raw files in the current working tree are authoritative.
2. Use CodeMemory first for orientation only when it is both available and permitted; treat it only as navigation context because it may be stale or incomplete.
3. Cite every finding and factual inventory assertion as path:line.
4. Classify conclusions as confirmed (direct current-file evidence), likely (bounded inference with evidence), or unknown (not established; name the missing evidence).
5. Verify referenced paths, command routes, agents, prompts, skills, instruction links, and allowlists directly.

Normalization and mirror logic

Normalize each asset as: ecosystem, kind, logical name, path, purpose, trigger, owner/dispatcher, expected output, mutation boundary, references, and relevance evidence.

Match mirrors by logical name and role, not by byte equality. Compare purpose, trigger, scope, inputs, output contract, evidence standard, and mutation boundary. Necessary OpenCode/Claude tool syntax, frontmatter, model, path, or dispatch translations are not defects when evidenced. Record evidenced translations under Deliberate differences; do not classify them as defects without contrary evidence.

Classification definitions

• Parity: mirrored artifacts preserve the same audit contract, scope, evidence standard, output, and mutation boundary, allowing evidenced harness translations.
• Duplicate mirror: one logical artifact intentionally exists in both ecosystems; not a defect by itself.
• Semantic overlap: distinct artifacts substantially cover the same purpose or trigger.
• Conflict: artifacts prescribe incompatible purpose, trigger, owner, output, scope, or mutation boundary for the same situation.
• Stale reference: a repository-controlled reference targets something missing, renamed, incompatible, or no longer routed.
• Coherence gap: a required purpose, trigger, owner, output, mutation boundary, route, or evidence rule is absent, ambiguous, or inconsistent.
• Relevance gap: an artifact is unsupported by the selected relevance mode's current stack/conventions, or relevant stack/conventions lack a needed harness counterpart. This is not a product-source-quality judgment.

gc.platform relevance criteria

In gc-platform mode, use only current repository evidence for relevance: project markers and documented conventions for the Angular frontend, .NET backend, tests, API specifications, and harness rules. Do not infer relevance from user-home assets or inspect application implementation quality. In portable mode, evaluate whether the artifact remains coherent as a generic OpenCode/Claude harness asset.

Report output contract

Return exactly one Markdown document with no preamble or postamble. If a section has no items, write None evidenced.

# AI Ecosystem Audit

## Audit identity
| Field | Value |
| --- | --- |
| Audit date | <date> |
| Revision | <SHA or unknown> |
| Scope | <parsed scope> |
| Strict mode | <on/off> |
| Relevance mode | <gc-platform/portable> |
| Included roots | <roots> |
| Exclusions | <exclusions> |

## Summary
| Status | Count | Evidence |
| --- | ---: | --- |
| Confirmed findings | <n> | <path:line links> |
| Likely findings | <n> | <path:line links> |
| Unknowns | <n> | <path:line links or `none`> |
| Deliberate differences | <n> | <path:line links> |

## Parity inventory
| Logical artifact | OpenCode | Claude Code | Parity | Evidence |
| --- | --- | --- | --- | --- |

## Findings
| ID | Classification | Confidence | Type | Evidence | Impact | Recommended next action |
| --- | --- | --- | --- | --- | --- | --- |

## Purpose map
| Artifact | Purpose | Trigger | Owner/dispatcher | Output | Mutation boundary | Relevance | Evidence |
| --- | --- | --- | --- | --- | --- | --- | --- |

## Duplicate/collision register
| Artifacts | Classification | Shared purpose/trigger | Disposition | Evidence |
| --- | --- | --- | --- | --- |

## Deliberate differences
| Difference | Why deliberate or evidenced | Evidence |
| --- | --- | --- |

## Risks and unknowns
| Item | Classification | Missing evidence / resolution path | Evidence |
| --- | --- | --- | --- |

## Prioritized backlog
| Priority | Item | Suggested owner | Validation | Evidence |
| --- | --- | --- | --- | --- |