Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add eddiebelaval/squire --skill epistemic-matrixgit clone --depth 1 https://github.com/eddiebelaval/squireWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/eddiebelaval/squire/epistemic-matrix)<a href="https://agentmods.dev/skills/eddiebelaval/squire/epistemic-matrix"><img src="https://agentmods.dev/badge/skills/eddiebelaval/squire/epistemic-matrix/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/eddiebelaval/squire/epistemic-matrix"><img src="https://agentmods.dev/badge/skills/eddiebelaval/squire/epistemic-matrix.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00162 | $0.01599 |
| Opus 5 | $0.00081 | $0.00800 |
| Sonnet 5 | $0.00032 | $0.00320 |
| Haiku 4.5 | $0.00016 | $0.00160 |
Grade A, and why
epistemic-matrix scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Epistemic Matrix
Core Workflows
Workflow 1: Primary Action
- Analyze the input and context
- Validate prerequisites are met
- Execute the core operation
- Verify the output meets expectations
- Report results
Compile a Rumsfeld epistemic-matrix audit over a scope. Two of these already exist on disk and define the house style — read them before writing:
EPISTEMIC-MATRIX-2026-07-08.md(portfolio-wide, the canonical form)EPISTEMIC-MATRIX-MEMORY-2026-07-14.md(subsystem-scoped, shows a headline + addendum)
The one rule
The migration is the asset, not the four lists. Every incident becomes an unknown-unknown migrating to a known-known; the machinery that does that migration is what the matrix audits. A matrix that just sorts facts into four buckets has failed. End every matrix on the migration engine and ranked next moves.
Doctrine (what makes this ours, not generic Rumsfeld)
- Q1 facts carry evidence on disk. Every known known cites a file, a
file:line, a dated log, a commit, or a board. No claim without a receipt. This isfeedback_verify_before_reportingenforced structurally. If you cannot ground it, it is not a Q1 fact — it is a Q2 question. - Q4 is breeding grounds, not items. You cannot list unknown unknowns by definition. You name the grounds where surprises breed — and each named ground must already have produced at least one real surprise. Speculation without a prior incident is not a breeding ground; it is noise.
- Q3 is anchored to the J-space rule. "An unsigned lesson is functionally unconscious — you can't steer what you can't say." Q3 hunts things already true on disk that no loop reads: dirty trees, inert insights, stale-but-adjudicated bets, tools never run, unsigned lessons. This is the weakest quadrant in most systems and the highest-value one to work.
- Stale claims are the enemy. A written "fact" that has gone false is worse than an unknown. Flag known-knowns going stale as their own shelf, and treat any load-bearing doc last touched weeks ago as suspect until re-verified.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 130 lines · 162 tokens per session scan A c466daf87c16
epistemic-matrix is a skill published in the GitHub repository eddiebelaval/squire (21 stars, last pushed 27d ago), licensed MIT. It adds 162 tokens to every session and 1,599 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
tdd-cycle
Red-Green-Refactor test-driven development cycle.
mcp-builder
MCP server development guide — build, test, and deploy MCP servers.
CSOC Operations & Playbook Automation
SOC alert triage, incident playbook automation, escalation workflows, shift reporting, and SOC KPI tracking.
beevibe-team-mesh-negotiation
Multi-round negotiation protocol — covers both initiator and peer roles. Use when about to call negotiate(), when receiving a intent block as a peer, or when receiving an 'escalated' sentinel from a blocked respondnegotiate. Covers proposal crafting, counter-strategy, deadlock detection, when to accept early…
rag-retrieval
Retrieval-Augmented Generation patterns for grounded LLM responses. Use when building RAG pipelines, embedding documents, implementing hybrid search, contextual retrieval, HyDE, agentic RAG, multimodal RAG, query decomposition, reranking, or pgvector search.
launch-runbook
Plan and execute a launch runbook covering pre-launch verification, go-live procedures, DNS cutover, post-launch monitoring, and rollback procedures. Use this skill whenever the user is preparing to launch a website or product, planning a DNS cutover, building a go-live checklist, or executing a launch day. Triggers…