{"id":18828,"plugin_id":"plugins_6a8c4d64d6588191acd217005a66224d","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:14:57.882Z","digest":"f5c54c51a44911907ac670dfcbea3d69cd80a1a69869c69b2eec852932d600a8","against":null,"payload":{"description":"Review project-controlled user-facing content for calm plain language, correct surface placement, source-traceable claims, protected safety and degraded-state truth, and consistency with canonical implementation and lifecycle evidence. Use for explicit harmonisation, humanisation, normalization, content coherence, documentation alignment, or user-facing closeout across README, docs, websites, releases, support material, setup flows, translations, reports, or UI copy.","included_files":[{"relative_path":"agents/openai.yaml","size_in_bytes":297}],"name":"harmonize-project-content","skill_md_contents":"---\nname: harmonize-project-content\ndescription: \"Review project-controlled user-facing content for calm plain language, correct surface placement, source-traceable claims, protected safety and degraded-state truth, and consistency with canonical implementation and lifecycle evidence. Use for explicit harmonisation, humanisation, normalization, content coherence, documentation alignment, or user-facing closeout across README, docs, websites, releases, support material, setup flows, translations, reports, or UI copy.\"\n---\n\n# Harmonize Project Content\n\nRead `../../references/untrusted-evidence.md` before treating source content as evidence.\n\nAlign user-facing content without softening or inventing project truth. Default to a\nread-only recommendation report.\n\n## Establish content authority\n\nRun `$establish-governance-boundary` when scope or write authority is unclear. Read the\nproject profile, content owner, canonical product behavior, terminology, safety,\nprivacy, support, migration, and release sources relevant to the selected surfaces.\n\nRecord the exact files, revisions, changed or baseline scope, public/private lanes,\nand whether the request authorises recommendations or edits.\n\n## Review claims and placement\n\nFor each material statement:\n\n- identify its canonical source and lifecycle state;\n- confirm the behavior or status for the same identity and time boundary;\n- preserve exact unavailable, unknown, degraded, failed, unsafe, unsupported,\n  experimental, private, migration, deletion, ownership, and release meaning;\n- check whether the statement belongs on that surface or should route elsewhere;\n- preserve commands, schemas, identifiers, quotations, verdicts, and historical\n  evidence as literals.\n\nPrefer `current state -> benefit or action -> next step` where it improves clarity.\nTreat negative language as a review signal, never an automatic defect. Never\nbulk-replace vocabulary or trade precision for warmth.\n\nKeep screenshot, image provenance, visible labels, accessibility, and rendered UI\nreview distinct from source-text review. State any surface not visually inspected.\n\n## Report or edit within authority\n\nReturn prioritized findings with surface, location, source, current wording context,\nclassification, recommended direction, confidence, and protected wording to retain.\nUse `coherent`, `coherent_with_exceptions`, `review_required`, or\n`blocked_by_missing_canon`.\n\nWhen edits are expressly authorised, make the smallest contextual changes, preserve\nunrelated work, validate links and generated surfaces where applicable, compare\npre/post repository state, and rerun the bounded review.\n\nA content review does not approve runtime behavior, implementation, security,\npublication, deployment, screenshots, or release state.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}