Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By jongwony
Evaluate epistemic risks in AI operations by scanning for upstream deficits, materializing intents into tasks, classifying risks, autonomously executing low-risk actions like file edits and builds via bash and git, and gating high-risk steps like git pushes for human review.
npx claudepluginhub jongwony/epistemic-protocols --plugin prosocheWhen AI collaboration goes off-track, you redo everything. These protocols catch direction errors early — at the plan level, not the code level. Fix the direction, not the implementation.
Correcting a wrong direction at the plan level costs one conversation turn. At the code level, it costs hours of rework. These protocols insert structured checkpoints at decision points — intent, goal, context, perspective, execution — so you and AI stay aligned before committing to implementation.
Install all protocols and utilities:
curl -fsSL https://raw.githubusercontent.com/jongwony/epistemic-protocols/main/scripts/install.sh | bash
Then run /onboard — start with a quick recommendation based on your recent sessions, then optionally continue to guided learning with scenarios, trials, and quizzes.
Paste in Codex and run:
$skill-installer install these skills from jongwony/epistemic-protocols:
- prothesis/skills/frame
- syneidesis/skills/gap
- hermeneia/skills/clarify
- katalepsis/skills/grasp
- telos/skills/goal
- aitesis/skills/inquire
- horismos/skills/bound
- analogia/skills/ground
- prosoche/skills/attend
- epharmoge/skills/contextualize
- epistemic-cooperative/skills/onboard
Restart Codex, then start with $onboard.
onboard. /report, /dashboard, reflexion, write are not included.skill-installer pattern with one path.| Protocol | Command | When to use |
|---|---|---|
| Hermeneia | /clarify | AI keeps building what you didn't ask for |
| Telos | /goal | You know you want something but can't define success |
| Aitesis | /inquire | AI charges ahead without asking what it needs to know |
| Prothesis | /frame | Need to look at this from multiple angles, not sure which |
| Analogia | /ground | AI's advice sounds right in theory but unclear in your context |
| Syneidesis | /gap | About to commit but might be missing something |
| Prosoche | /attend | Want execution readiness checked and risky actions gated for your judgment |
| Epharmoge | /contextualize | AI's output is correct but doesn't fit your situation |
| Horismos | /bound | You need to define what you know vs what AI should figure out |
| Katalepsis | /grasp | AI made big changes and you need to actually understand them |
Concern clusters: Planning (/clarify, /goal, /inquire) · Analysis (/frame, /ground) · Decision (/gap) · Execution (/attend) · Verification (/contextualize) · Cross-cutting (/bound, /grasp)
| Plugin | Command | Purpose |
|---|---|---|
| Epistemic Cooperative | /onboard, /report, /dashboard | Protocol learning, usage analysis, and coverage dashboard |
Each protocol targets a specific point where AI collaboration can go wrong. For architecture details and design philosophy, see CLAUDE.md.
| Protocol | Greek | Meaning |
|---|---|---|
| Prothesis | πρόθεσις | Setting forth |
| Syneidesis | συνείδησις | Shared knowing |
| Hermeneia | ἑρμηνεία | Interpretation |
| Katalepsis | κατάληψις | Grasping, comprehension |
| Telos | τέλος | End, purpose |
| Horismos | ὁρισμός | A bounding |
| Aitesis | αἴτησις | Request, inquiry |
| Analogia | ἀναλογία | Proportion |
| Prosoche | προσοχή | Attention |
| Epharmoge | ἐφαρμογή | Application, fitting |
MIT
Uses power tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Infer context insufficiency before execution — /inquire (αἴτησις: a requesting)
The epistemic-posture layer for AI coding agents. Ships the Reasoning Surface protocol, named failure-mode counters, operator profile schema, and workflow loop as Claude Code skills, agents, and hooks. Posture first. Kernel outlives the tooling.
YES.md — PUA says NO, YES says YES. 6-layer AI governance: format → trigger → hooks → anti-slack → gates → memory. Makes AI do things RIGHT with encouragement, not pressure. Available in English, 中文, 日本語.
Achieve flow state safely with Claude Code. Auto-approves routine work, gates risky actions, hard-blocks dangerous patterns. Dual enforcement (skill + hooks), token cap for cost governance, full audit trail. Zero dependencies.
AI agent skill pipeline: plan-interview, intent-framed-agent, context-surfing, simplify-and-harden, self-improvement, and agent-teams-simplify-and-harden. Prevents scope drift, context degradation, rough code, and repeated mistakes.
Core safety skills for AI-assisted development: Four Laws, Three Strikes, production-first, scope validation, and environment separation
Define epistemic boundaries per decision — /bound (ὁρισμός: a bounding)
Detect application-context mismatch after execution — /contextualize (ἐφαρμογή: an application)
Achieve certain comprehension of AI work — /grasp (κατάληψις: a grasping firmly)
Infer context insufficiency before execution — /inquire (αἴτησις: a requesting)
/catalog for protocol handbook, /onboard for quick recommendation + protocol learning, /report for Growth Map (usage analysis + insights integration), /dashboard for coverage dashboard, /compose for protocol composition authoring, /sophia for philosopher match (behavioral dimension analysis), /curses for strength-shadow analysis and attitude recommendations
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claim