jongwony/epistemic-protocols

Aitesis Protocol scans for missing context before execution and asks prioritized questions to resolve uncertainties. Use it when an AI needs to separate factual, coherence, and relevance gaps before proceeding.

Analogia Protocol validates structural mappings between abstract and concrete domains. It surfaces concrete instantiations when a mapping is uncertain and supports user grounding, repair, or withdrawal.

Contextualize detects when an output is technically correct but does not fit the current application context. Use it after execution to surface applicability mismatches for review, routing, or adaptation.

Elenchus Protocol vets working context before action by challenging claims, checking provenance, and surfacing counterarguments. Use it when a context may be suspect and needs to be resolved into a vetted context.

Elicit reverse-traces decision coordinates from an ambiguous intent using the user’s codebase, rules, session history, and environment. Use it when a request is underspecified and you need to surface questions, parked coordinates, and a resolved endpoint.

Encapsulation Audit reviews plugin SKILL.md prose and plugin description metadata for self-contained language. Use it when checking for contributor-knowledge assumptions or banned-reference rephrasings that evade deterministic encapsulation checks.

Forge forms a prompt artifact by grounding under-determined intent in an authoritative reference document. It surfaces required contract coordinates and projects them into a follow-up prompt or standing custom-skill recipe.

Formal Review runs a fixed Category Theory, Type Theory, and Operational Semantics pass over a PR diff and posts the surviving findings as one consolidated comment. Use it for repository changes that need a structured formal review.

Gate Check freezes a drafted option set and sends it to an independent advisor for a genuine, collapsed, or malformed verdict. It then verifies the cited grounds against the substrate before the result is returned.

Goal Research delegates a research question to Codex CLI, which scopes the task with `/goal` and uses Aitesis Tavily searches for external verification. Use it when you need cited fact-finding or to check claims against outside sources.

Heuresis Protocol widens an underexpanded idea field by generating diverse candidates in parallel without ranking or eliminating them. Use it when a topic is blank or has converged too early and you need more possibilities before choosing.

Onboard recommends an epistemic protocol from recent sessions, then can guide the user through scenario, trial, and quiz. Use it for new-user onboarding, re-onboarding, or protocol discovery.

Placement Audit routes each clause of LLM-facing instruction prose to a destination and flags any surface that is not fully audited. Use it when deciding where a clause belongs, auditing an instruction file, or tracing why a prompt keeps growing.

Probe surfaces multiple candidate epistemic deficits for a vague situation and asks the user to recognize the best fit. Use it when something feels off but the right protocol is unclear.

Protocol Verification checks epistemic protocol definitions for consistency before commit. Use it to run static checks and expert review when validating definitions, soundness, or pre-commit quality.

Realign fuses the project guide's current direction, external work signals, and the user's present understanding into one direction line. Use it when the guide no longer matches what the project is doing and a traced update is needed.

Recollect resolves vague recall by scanning contextual clues and presenting grounded excerpts for the user to recognize. Use it when you remember a discussion, topic, or concept but cannot name it yet.

Reduced-Space Test validates an uncertain claim in a bounded stand-in space, then reports what holds, fails, or stays inconclusive under those conditions. It also carries the untested remainder forward for follow-up.

Release creates a CalVer-tagged release and prepends a hand-written narrative to generated notes. Use it when you want to publish a repo release with CI-built artifacts and a human summary.

Review Loop runs a convergence-paced review-resolve cycle over a code change and its governing surfaces. It verifies findings against the codebase and the review base, then re-reviews until each finding is disposed of or left as residual.

Steer audits a target session for calibration drift, groups evidence by cluster, and drafts an updated project-profile rule for user approval. It also records settled directions in the project guide when a resolution has become a recurring convention.

Triage forms focused work units from GitHub issues by grouping related reports and fusing them with the project northstar. It is used when you need to route issue clusters into handoff records for another session to execute.

Type Realization measures whether a named protocol’s formal contract is realized during an actual run. Use it to test runtime behavior like phase transitions, gates, and stop/proceed outcomes rather than static file quality.

White Bear Audit reviews LLM-facing prose for unnecessary mentions of forbidden acts, retired paths, or rejected alternatives. Use it when checking instruction files, prompts, or output-style text for negative framing that may compete with the intended path.

Zero Shot Audit reviews LLM-facing prose for few-shot anchoring and example lists that narrow a principle’s application. Use it to spot read-only findings in skill, agent, or output-style instructions.