jongwony-epistemic-protocols

Plugin jongwony-epistemic-protocols from marketplace cascade-ai-ml-engineering in joshuarweaver/cascade-ai-ml-engineering. Commands read from the repo's manifest — 1 month ago.

Skills (31)

• encapsulation

  This skill should be used when the user asks to \"audit plugin encapsulation\", \"check self-containment semantics\", \"find contributor-knowledge assumptions\", or invokes /encapsulation. Reviews LLM-facing prose in this project's plugin SKILL.md files and plugin description metadata for Plugin Enc

• formal-review

  This skill should be used when the user asks to \"formal review\", \"formal lens review\", or invokes /formal-review. A fixed-lens PR review for this repository's formally-structured protocol changes: it pins a Category Theory / Type Theory / Operational Semantics lens panel over only the files chan

• outcome

  This skill should be used when the user asks to "run the outcome eval", "paired bare vs protocol", "which decisions did the protocol surface", "count what the AI asked or presented", "does /inquire surface what the request left out", or wants to see, from transcripts, which decisions reached the use

• realize

  This skill should be used when the user asks to "run the eval", "test whether the protocol actually works at runtime", "check type realization", "measure protocol fulfillment", "run the type-realization suite", "does the gate actually stop", or wants runtime evidence that a protocol's formal transit

• rederive

  This skill should be used when choi names a core protocol whose Lean contract is to be re-derived, or invokes /rederive. Runs the per-protocol flow: gather the pre-understanding, derive the protocol's epistemic solution and fuse it with choi's horizon, then read the current contract against it, thro

• release

  Release for this repo — CalVer tag + narrative prepend.

• verify

  This skill should be used when the user asks to "verify protocols", "check consistency before commit", "validate definitions", "run pre-commit checks", "verify soundness", or wants to ensure epistemic protocol quality. Invoke explicitly with /verify for pre-commit validation.

• inquire

  A task rests on missing context or unchecked assumptions: collect every piece of context the AI can reach on its own, then hand back what it cannot reach as the user's own unknown.

• ground

  A framework, analogy, or earlier design is carried to a case, or an abstraction tested against its own cases, and what it supports is unclear: back each conclusion with cited evidence, not assent.

• recollect

  The user vaguely recalls something discussed before but cannot name it — one session, or a line of work, topic, or settled concept across several: find it in past records to recognize.

• sublate

  Context about to be acted on may no longer hold — stale, weakly sourced, or contradicted: challenge each claim the action leans on, and the user judges what stands before acting.

• contextualize

  A result may not fit where it lands: check every place it reaches and every intent it was meant to carry, show every misfit and omission on one sheet, and carry out the fixes the user settles.

• gate-check

  Advisor-checked decision gates. Freezes a drafted option set, has an independent external advisor rule it genuine / collapsed / malformed, then verifies the cited grounds before it reaches the user.

• goal-research

  Delegate factual research to a Codex CLI session that uses Codex's builtin `goal` command and invokes Aitesis (`$inquire`) for Tavily-based external verification. User-invoked via /goal-research.

• onboard

  Quick protocol recommendation from recent sessions, or quest-based learning through scenario, trial, and quiz.

• probe

  Deficit Recognition Probe — surface multiple deficit hypotheses for the user's current situation and route by user-constituted recognition (fit review, not protocol scoring).

• realign

  Project guide direction line via three-horizon fusion. Invoke with /realign.

• reduced-space-test

  Scoped empirical validation. Decomposes a target↔surrogate equivalence claim into facets, bounds a test space, captures evidence inside it, and carries the untested complement forward.

• review-loop

  Convergence-paced review-resolve loop over a change and its governing surfaces. Verifies each finding against the codebase and the base it is measured from, then re-reviews until each is disposed of.

• white-bear

  Use when the user asks to \"check white bear\", \"audit prohibitions\", \"find negative framing\", or invokes /white-bear. Read-only audit: unnecessary competing-target mentions in LLM-facing prose.

• zero-shot

  Use when the user asks to \"check zero-shot\", \"audit few-shot anchoring\", \"find example anchoring\", or invokes /zero-shot. Read-only audit of LLM-facing prose: principle over anchoring examples.

• elicit

  The user knows roughly what they want but not which decisions it turns on: trace them from their own material (code, rules, past sessions) and the domain's usual decisions; ask until it settles.

• ideate

  The idea field is empty, narrowed too early, or thin in places: widen it in width and depth before any candidate is chosen; reads the user's words and what they name.

• bound

  The user cannot yet see what a task needs them to decide, or which decisions to keep or entrust: map the whole task first, then open each decision to the depth needed.

• conduct

  The work needs several lines of thinking, and their order, independence, combination, stopping point, or where results go is not obvious: settle the method before starting.

• sketch

  Something has to be made that the user cannot describe but would recognize on sight: mark what does not fit on concrete sketches, revising until a form is recognized.

• grasp

  Something in play — code, a document, a result, quotable in context — needs to be actually understood; the user can't follow it yet or nods along unsure: verify understanding step by step.

• apportion

  One goal is about to be handed to an autonomous run: cut it into units that each fit one stretch of the run and carry their own conditions for when they are done.

• induce

  Several concrete cases at hand seem to share something no one has named yet: pin down what they have in common, checked against the goal the name serves.

• preview

  The user is about to commit to one of several directions that cannot be judged from their descriptions: build throwaway probes showing where they diverge, and decide on what is seen.

• route

  Route the accumulated session context to the loaded core protocol whose deficit it shows — /route. Invokes the dominant match, nudges when several fit, silent when none, monitors while a gate holds.

Similar plugins