Skip to content
cccskills
FreePublished on GitHub

cognitive-compile

Use when source material must be transformed into a compact, evidence-aware model for learning, decisions, or an Obsidian concept note.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141 stars(GitHub 141 + cccskills 0)
Works withObsidian

Run score Not tested

See how to install

Files included (1)

  • SKILL.md3.9 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Cognitive Compile

<skill_contract> <input>Source-grounded material, original question, learning or decision goal, audience, and destination.</input> <output>An eight-part evidence-aware mental model separating facts, mechanisms, conflicts, hypotheses, decisions, and action.</output> <done>Each material claim retains provenance and uncertainty, and one bounded next action answers the original question.</done> <non_goals>Inventing missing claims, erasing contradictions or provenance, or treating interpretation as source fact.</non_goals>

Compile information into a decision-ready mental model. Compression must preserve provenance, uncertainty, contradiction, and the original question.

Usage Template

Provide: source material, original question, intended decision or learning goal, and preferred destination. Optional: existing wiki concepts to compare.

Workflow

<intake>

Identify the source boundary, original question, audience, and required fidelity. Label direct source claims, external facts, and analyst inferences before synthesis.

</intake>

<unknowns_gate>

If the source is inaccessible, truncated, or missing the material needed to answer the question, return INSUFFICIENT_EVIDENCE. Ask one probe only when its answer changes the compile structure; otherwise proceed with explicit unknowns.

</unknowns_gate>

<execute>

Produce eight compact sections:

  1. Original question — the problem the material helps answer.
  2. Key facts — source-grounded claims with locators when available.
  3. Concepts and entities — stable terms worth linking.
  4. Patterns — mechanisms, feedback loops, or recurring structures.
  5. Conflicts — contradictions with the source, prior knowledge, or internal claims.
  6. Hypotheses — falsifiable interpretations, explicitly labeled.
  7. Decision support — options, tradeoffs, and what the evidence changes.
  8. Action — one bounded next test, note, or decision.

Prefer one durable concept note over several thin notes. Preserve source links or block references; do not rewrite immutable source notes.

</execute> <evaluate>

Verify that every material conclusion traces to evidence or is labeled inference; the original question is answered; contradictions remain visible; and the action follows from the model. Remove summaries that merely restate the source without changing understanding.

</evaluate>

Failure Protocol

  • NEEDS_INPUT: the intended question or audience is absent and materially changes the output.
  • INSUFFICIENT_EVIDENCE: the source cannot support the requested compile.
  • VERIFY_FAILED: a conclusion lacks traceable evidence; downgrade it to hypothesis or remove it.

Output Contract

Return status, result (the eight-section compile), evidence (source locators), unknowns, and next_action.

Edge Cases

  • Two sources disagree: preserve both claims, compare evidence quality, and state what observation could resolve the conflict.
  • The source is mostly opinion: compile mechanisms and assumptions, but do not promote assertions into facts.

Success Metrics

  • The result answers one explicit question with traceable evidence.
  • Facts, conflicts, and hypotheses are distinguishable at a glance.
  • At least one decision or test becomes clearer.

Quality Gates

  • Source boundary and provenance are explicit.
  • All eight sections are present or marked not applicable with reason.
  • No hypothesis is phrased as established fact.
  • The next action is bounded and verifiable.

</skill_contract>

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Use when work has genuinely independent streams or distinct builder, evaluator, domain, and integration roles that require bounded multi-agent command scaled from 5 to 100+ agents.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

Use when designing or refactoring a model-native engineering workflow with bounded autonomy, probes, custom evaluation, durable state, and verified write-back.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

Use when property-service operations need an AI plus ontology plus DMAIC design for work orders, dispatch, quotes, evidence, CTQ metrics, and control dashboards.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

Use when a personal or team operating system needs a bounded redesign using Four-C, closed-loop controls, 70/30 allocation, 3B creativity, experiments, and prediction-error learning.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

Use when a goal must be converted into a repeatable behavior, cue, SOP, review cadence, and identity-aligned reinforcement.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

Use when a long-running agent task needs context budgeting, checkpointing, compaction, retrieval, or capability-based model routing.

The English summary is being prepared. Showing the original description.

Mark393295827/third-brain-v7-skills141Updated Sep 19, 2026

See all skills by Mark393295827

Report a problem with this skill