Borrowing it
Nothing to install: this file belongs to ljtn/epiq. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ljtn/epiq/main/.claude/skills/epiq-architecture/SKILL.mdgit clone --depth 1 https://github.com/ljtn/epiqWrote 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/ljtn/epiq/epiq-architecture)<a href="https://agentmods.dev/skills/ljtn/epiq/epiq-architecture"><img src="https://agentmods.dev/badge/skills/ljtn/epiq/epiq-architecture/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/ljtn/epiq/epiq-architecture"><img src="https://agentmods.dev/badge/skills/ljtn/epiq/epiq-architecture.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00055 | $0.01206 |
| Opus 5 | $0.00028 | $0.00603 |
| Sonnet 5 | $0.00011 | $0.00241 |
| Haiku 4.5 | $0.00006 | $0.00121 |
Grade A, and why
epiq-architecture 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 6d 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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The event log is a CRDT
There is no server and no shared clock. Every actor appends to its own JSONL log on the state branch, git merges them, and each machine derives the same board from the same set. Everything below exists to keep that derivation a pure function of the event set.
The model
id = [ulid, refId].refIdis the causal parent: the tail of the causal order this actor last saw (getEdgeRef). Concurrent writers legitimately share a parent.- Order is derived, never stored.
getSortedEventsrebuilds the forest fromrefId, sorts concurrent siblings by ULID, walks depth-first, and dedupes by id. File line order is not load-bearing. - The ULID is a hybrid logical clock, not a timestamp:
getNextId(Math.max(Date.now(), decodeTime(edge) + 1)). The wall clock is only a lower bound; causality forces monotonicity. - Actor id comes from the file name; the display name does not.
persistcallsstripActor, so the payload carries neither. The id is parsed from the log's file name; the name is resolved from the contributor registry, whichcreate.contributor/rename.contributorbuild. The name segment in the file name is a sanitized storage key, not a name of record. - Same event set ⇒ same order ⇒ same board. That is the whole contract.
Invariants
- Derive order from
refIdplus the ULID tiebreak. Nothing else. - Ids are permanent. Tombstone instead:
tombstoneNodemarks the node and its descendantsisDeleted;tombstoneContributorclears the display name but keeps the record so assignments referencing the id still resolve. - Replay is total. An event that lost a race or names state this replay never applied is a
materializeSkip(ConvergenceFail) and is skipped. The same precondition on a live write is a genuine failure — there the precondition is the answer the caller asked for. - Unreadable stays ordered. The envelope (
v,id) parses even when the payload cannot, so an event from a newer build keeps its place in the chain. - Append only. One id always means one byte sequence. This is what makes
*.jsonl merge=unionsafe. - Time travel cuts causally.
splitEventsAtTimemarks a child unapplied when its parent is unapplied, whatever its own timestamp says.
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.
- 6d ago Changed · +9 lines 56f7e4ebb034
- 10d ago First seen · 51 lines · 55 tokens per session scan A 7776b12c0a6a
epiq-architecture is a skill published in the GitHub repository ljtn/epiq (366 stars, last pushed yesterday), licensed MIT. It adds 55 tokens to every session and 1,206 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
conventional-commit
Workflow for generating conventional commit messages following the Conventional Commits specification. MUST be invoked every time a commit is created. Guides construction of standardized commit messages with correct type, scope, description, body, and footer.
resolve-conflicts
Use this skill immediately when the user mentions merge conflicts that need to be resolved. Do not attempt to resolve conflicts directly - invoke this skill first. This skill specializes in providing a structured framework for merging imports, tests, lock files (regeneration), configuration files, and handling…
github-pr-description
Generate and create pull request descriptions automatically using GitHub CLI. Use when the user asks to create a PR, generate a PR description, make a pull request, or submit changes for review. Analyzes git diff and commit history to create comprehensive, meaningful PR descriptions that explain what changed, why it…
python-release
Handle Python SDK release, build, bump, packaging metadata, PyPI client pin, uv.lock, nox/build workflow, and publish verification changes. Use for Python release process work or dependency pin bumps; do not use for ordinary Python feature implementation.
write-release-notes
Generate engaging, high-energy release notes for a given version tag. Fetches the release from GitHub, retrieves every linked PR's title and description, then synthesizes all changes into a polished, user-facing release note with an enthusiastic tone. Use when the user asks to write, generate, or create release notes…
create-github-issue
Create GitHub issues using GitHub CLI with support for templates, labels, assignees, milestones, and draft issues. Use when the user asks to create a GitHub issue, file a bug report, submit a feature request, or open an issue in a GitHub repository.