From superpipelines
Enforces standards-compliance auditing of pipeline bundles with FULL/DELTA/SCOPE-WIDE modes, compliance matrix dispatch, severity classification, fix routing, and model-tier resolution criteria.
How this skill is triggered — by the user, by Claude, or both
Slash command
/superpipelines:pipeline-auditor-protocolThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
<glossary>
<operating_modes>
| Mode | Trigger | Scope |
|---|---|---|
| FULL | Direct invocation or first audit. | Entire pipeline bundle: all agents, skills, and topology. |
| DELTA | Triggered by mutation commands. | Changed files plus immediate neighbors and the entry skill. |
| SCOPE-WIDE | audit-steps --all command. | FULL audit across all registered pipelines in all scope roots. |
| </operating_modes> |
entry.md under .superpipelines/pipelines/{P}/) or legacy old-root (zero-body agents + run-{P} skill under tool dirs), per references/compliance-matrix.md § "Pipeline layout & criterion applicability". Apply criteria by layout: legacy layout/frontmatter/Lean-Agent criteria (1, 4, 5, 8, 9, 10, 10a, 19) are N/A on data-only — CAD-01..CAD-05 govern there instead. Never FAIL a conformant CAD against a legacy-only criterion.references/compliance-matrix.md (criteria 1–25 including 10a, resolver consolidation criteria PR-01..PR-05, PR-07..PR-10 per ADR-0001 and ADR-0002, plus canonical agent-def criteria CAD-01..CAD-05), honoring the layout applicability table. Criterion 25 (frontmatter ↔ protocol capability coherence) flags any agent/protocol pair whose primary action assumes a tool the agent forbids; distinguish a blocked primary path (SEV-1) from a documented tier-conditional fallback (PASS). CAD-01..CAD-05 validate tool-neutral canonical agent defs under superpipelines/pipelines/{P}/agents/ (capability-intent consistency, relative io_contract paths, version stamping, reviewer write-deny).references/topology-rules.md.references/severity-classification.md and select fixes from references/fix-templates.md.disallowedTools: Write, Edit, Bash) and NEVER writes the report file or mutates registry.json. Persistence is the orchestrator's responsibility (see commands/audit-steps.md REPORTING).references/audit-report-template.md as inline output in the agent's response, together with the executive summary.audit/latest.md path and the last_audit timestamp value to record.pipeline-architect with the remediation plan.| ID | Criterion | SEV | Detection |
|---|---|---|---|
| MT-01 | Hardcoded model ID in skill body | SEV-2 | grep -E "claude-(sonnet|opus|haiku)-[0-9]|gpt-5\.[0-9]|gemini-3\." skills/**/SKILL.md skills/**/references/*.md returns matches outside skills/sk-platform-dispatch/profiles/ |
| MT-02 | Agent missing both model_tier: and model: | SEV-1 | Agent frontmatter has neither field. Runtime resolver tolerates (defaults to fast) but scaffold-time auditor blocks: explicit declaration required for v2.0+ agents. |
| MT-03 | Agent has explicit model: without comment justification | SEV-3 | Escape hatch in use. Surface to reviewer; do not block. Distinguish by plugin_version: if absent or < 2.0.0 → v1 legacy (Phase 0.45 migrates automatically); if >= 2.0.0 → intentional v2 escape hatch (advisory only). |
| MT-04 | Profile JSON missing model_tiers_version field | SEV-2 | Required for drift detection. |
| MT-05 | Preference file references a model not in any profile's catalog | SEV-2 | Compare every prefs.platforms[*].tiers[*] value against the union of all profiles' model_tiers[*].model. Mismatch likely typo or stale ID. |
| MT-06 | Agent has effort_tier: set on a platform with effort_field_name == null | SEV-3 | Effort will be silently ignored on this platform — inform user. Detection requires knowing the source/runtime tier. |
platform_profile.model_tiers[tier].model_tier: to the agent (architect should have done this in Phase 4).model: documenting WHY the escape hatch is needed."model_tiers_version": "YYYY-MM-DD" to the profile./superpipelines:change-models Mode F (catalog refresh) to reconcile.effort_tier: or accept that it's a no-op on this platform.
${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/compliance-matrix.md${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/canonical-agent-def.md${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/topology-rules.md${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/severity-classification.md${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/audit-report-template.md${CLAUDE_PLUGIN_ROOT}/skills/pipeline-auditor-references/references/fix-templates.mdnpx claudepluginhub gustavo-meilus/superpipelines --plugin superpipelinesProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.