FactHarbor

Keep repository documentation lean, clear, and readable. Use automatically before adding a new document, substantially expanding or rewriting Markdown/xWiki content, adding explanatory/background sections, adding FAQs/glossaries/templates, introducing repeated framing, or reviewing a documentation diff for clutter, filler, duplication, over-broad prose, stale links, or readability problems.

MUST be used automatically for any FactHarbor bugfix, regression fix, failing test/build fix, failed-validation recovery, or defect repair — BEFORE editing code. Decides whether the best fix is reverting or amending a previous change, quarantining/deleting obsolete code, or adding new code. Also use when about to add a fallback, flag, retry, helper, guard, diagnostic, or parallel mechanism — in response to a failure or proactively (even with no failing test) — or when reviewing a diff for workaround stacking / additive repair drift.

Analyze local job reports from the most recent build/commit (HEAD by default), detect quality issues, and propose AGENTS.md-compliant fixes. Scopes to HEAD by default, or to specific job URLs / job IDs / commit / input slug passed as argument. Uses RAG (static register + handoffs + machine-readable benchmark expectations, plus dynamic follow-up scans) and runs an adaptive multi-agent debate before emitting recommendations. Complements `/prompt-diagnosis` (which is narrower — prompt provenance only) by covering evidence, boundaries, verdict reasoning, and warning severity end to end.

Structured adversarial debate on any proposition. Runs an intake/structural audit first, then spawns pipeline-mirroring roles (Advocate, Challenger, Reconciler) with optional Consistency Probes and Validator. Use when a decision needs adversarial pressure before committing — root-cause attribution, architecture choices, fix selection, design tradeoffs. Other skills invoke /debate for their adversarial synthesis steps. Adapts role count and model tiers to proposition complexity.

Consolidate all documentation in Docs/WIP/. Investigates implementation status of WIP topics using helper agents when the tool supports delegation, ensures all open items are well-described and present in the backlog, updates and cleans up backlog and status files, and moves historical/completed content to the Archive. Use this skill whenever the user says "consolidate WIP", "clean up WIP", "tidy WIP docs", "audit WIP", or asks to update the backlog from WIP — even if phrased differently. Note: this skill covers Docs/WIP/ only. To clean up the entire Docs/ tree use docs-update.

Analyze debug-analyzer.log and run tests to identify pipeline issues after a change. Standard post-change check per Captain's protocol.

Deep analysis of the FactHarbor CB pipeline. Use for debugging, architecture decisions, evidence quality issues, or reviewing changes that span multiple pipeline stages.

RAG-augmented diagnosis of LLM prompting issues. Correlates pipeline failures with runtime prompt provenance (`promptContentHash`) plus execution commit context (`executedWebGitCommitHash`). Detects report-specific failures, systemic deficiencies, and prompt rollout drift. Use when a pipeline output looks wrong and the cause may be in how the LLM was instructed or propagated.

Run FactHarbor post-change validation on benchmark families and analyze results. CAUTION — submits real analysis jobs. Use after pipeline or prompt changes to verify no regressions.

Specialized agent for handling role activation, task completion logging, shared workflow-skill discovery, and following project-specific protocols (Role Activation Protocol, Agent Exchange Protocol). Use when switching roles, finishing non-trivial tasks, or needing to follow FactHarbor's Senior Engineer engineering standards.

Static quality audit of LLM prompts in apps/web/prompts/. Scores each prompt against a 9-criterion rubric (rule compliance, efficiency, effectiveness, un-ambiguity, generic hygiene, multilingual robustness, bias/neutrality, output schema alignment, failure-mode coverage) and proposes AGENTS.md-compliant fixes. Static linter-style analysis — no runtime data, no job runs, no file writes. Complements /audit (broader prompt+code) and /prompt-diagnosis (runtime provenance).

Clean up all documentation files (.md and .xWiki) across the Docs tree. For living documentation outside WIP (especially .xWiki specs, architecture, and reference pages), the primary action is to UPDATE content to match the current codebase and decisions — not archive it. Only content that is clearly and entirely obsolete is moved to the Archive. Partially valid content is always updated in place, never lost. Diagrams (Mermaid) are actively verified and updated to reflect current architecture. For WIP files, the skill archives completed/historical items and syncs the backlog. Use this skill whenever the user asks to "clean up docs", "update documentation", "tidy docs", "remove obsolete content", "archive stale pages", or anything similar across .md or .xWiki files. Note: for WIP-only consolidation with helper-agent investigation, use wip-update instead.

Generate a properly-formatted FactHarbor handoff document at end of a task. Captures completed work, decisions, warnings, and next steps for the next agent or session.

Full FactHarbor quality audit — all prompts and key pipeline stages. Use when investigating analysis quality regression, before a release phase, or after major pipeline changes.

Explains code with visual diagrams and analogies. Use when explaining how code works, teaching about a codebase, or when the user asks "how does this work?"