agentsclimarketplace

Epistemic mapping

Skill r-irbe/proof-skills/skills/epistemic-mapping

APM-installable agent skills for Lean 4 and Mathlib4 — proof tactics, math domains, review and research workflows, generic tooling.

Install
npx -y skills add r-irbe/proof-skills --skill epistemic-mapping

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

USE FOR: mapping a domain's concept lattice, identifying load-bearing assumptions, surfacing implicit ontologies, drafting an epistemic terrain map. DO NOT USE FOR: producing formal proofs (use @lean-proof), gathering primary research data (use @applied-intelligence-analysis). TRIGGERS: concept-map, ontology, terrain, epistemic-survey.

SKILL.md

2.6 KB, as published. Nobody here has run it

epistemic-mapping

Routing

  • USE FOR: mapping the concept lattice of an unfamiliar domain; identifying load-bearing assumptions; surfacing implicit ontologies; drafting an epistemic terrain map to seed downstream formalization.
  • DO NOT USE FOR: producing formal proofs (use @lean-proof); gathering primary research data (use @applied-intelligence-analysis).
  • TRIGGERS: concept-map, ontology, terrain, epistemic-survey.

Behavioural rules (G-*)

  • G-1 (MUST): The map MUST distinguish observed concepts from inferred ones. [Trace: AC-01]
  • G-2 (SHOULD): The map SHOULD enumerate at least three competing framings before settling. [Trace: AC-02]
  • G-3 (MUST NOT): The skill MUST NOT collapse contested distinctions into a single label without an explicit caveat. [Trace: AC-03]
  • G-4 (SHOULD): The skill SHOULD cite at least one anchor source per top-level node. [Trace: AC-04]
  • G-5 (MUST): On confidence < 80 % the skill MUST mark the node provisional. [Trace: AC-05]

Workflow

  1. Discover [discover] — read brief, locate prior maps, list candidate concepts.
  2. Plan [discover] — choose framing set; STOP if < 3 framings.
  3. Draft [execute] — write the terrain map with provisional nodes flagged.
  4. Verify [validate] — cross-check anchor citations; max 3 attempts then escalate.
  5. Persist [persist] — commit map, update state tracker, tick tasks.md.

Recovery & STOP

  • Anchor citation missing ×3 → STOP, escalate.
  • Single-framing collapse detected → immediate STOP, re-anchor.
  • Confidence < 80 % on top-level structure → STOP, ask.

Handoffs

  • Predecessors / successors: see FM handoffs (grammar from ADR-0080).
  • Source spec: specs/epistemic/mapping/requirements.md.
  • Related ADRs: ADR-0076, ADR-0080.

Common failure modes

AI agents commonly: pick the first framing; merge contested terms; omit citations for "obvious" nodes. Full registry: GUARDRAILS.md.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.