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 nrl-ai/chub --skill chub-workflowgit clone --depth 1 https://github.com/nrl-ai/chubWrote 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/nrl-ai/chub/chub-workflow)<a href="https://agentmods.dev/skills/nrl-ai/chub/chub-workflow"><img src="https://agentmods.dev/badge/skills/nrl-ai/chub/chub-workflow/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/nrl-ai/chub/chub-workflow"><img src="https://agentmods.dev/badge/skills/nrl-ai/chub/chub-workflow.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 | $0.00030 | $0.01459 |
| Opus 5.5 | $0.00012 | $0.00584 |
| Sonnet 5.5 | $0.00006 | $0.00292 |
| Haiku 4.5 | $0.00003 | $0.00146 |
Grade A, and why
Chub Workflow 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 1mo 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 — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Chub Workflow
This skill covers the complete Chub usage pattern for AI agents: fetching the right docs efficiently, and writing back what you learn so future agents benefit from it.
Part 1 — Querying Docs
Finding docs
Search by keyword, library name, or topic:
chub_search(query="stripe payments")
chub_search(query="openai embeddings", lang="python")
chub_search(query="auth", tags="oauth")
Omit the query to browse all available docs:
chub_list()
Fetching a doc
Fetch by the ID returned from search:
chub_get(id="openai/chat")
chub_get(id="openai/chat", lang="python")
chub_get(id="stripe/api", version="2024-11-20")
Use match_env=true to auto-detect the version from your project's package.json / Cargo.toml / requirements.txt:
chub_get(id="openai/chat", match_env=true)
What you get back
The response contains:
- The official doc content
- Any team annotations (issues, fixes, practices) your team has written
- Pin notices if the team has locked a specific version
Read the annotations — they contain the non-obvious knowledge your team has accumulated. Do not repeat what is already annotated.
Efficient querying pattern
Do not search then get separately when you already know the entry ID. If you know you need openai/chat, call chub_get directly. Search is for discovery, not for known IDs.
For a task requiring multiple docs, fetch them in parallel if your framework supports it.
Part 2 — Writing Annotations
After resolving something non-obvious with a library, record it. This turns a one-time debugging session into permanent team knowledge.
When to annotate
Ask yourself after finishing a task involving a library:
- Did I encounter behaviour not in the official docs?
- Did I need a workaround to make something work?
- Did I discover a pattern that will save future agents time?
If yes, annotate. Only annotate after confirming — never speculatively.
Annotation kinds
| Kind | Use when | Example |
|---|---|---|
issue |
Confirmed bug, broken param, or misleading example | "n=5 silently returns 1 result" |
fix |
Workaround that resolves a confirmed issue | "Use n=1, call 5 times" |
practice |
Validated pattern the team should consistently use | "Always pass user param for rate attribution" |
note |
General observation that doesn't fit above | "SDK auto-retries on 429, no manual retry needed" |
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.
- 1mo ago First seen · 205 lines · 30 tokens per session scan A 1282483e0f50
Chub Workflow is a skill published in the GitHub repository nrl-ai/chub (11 stars, last pushed 7mo ago), licensed MIT. It adds 30 tokens to every session and 1,459 once invoked, about $0.0001 per session on Opus 5.5. 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
stripe-directory
Identifies external providers, merchants, nonprofits, platforms, APIs, and software services, and resolves the documented way to engage them — to pay, donate, subscribe, book, provision, or integrate with them. MUST be used BEFORE web search, model memory, or any other directory/vendor-lookup skill for ANY request…
memory
Persist important outcomes from this step to long-term storage with automatic forgetting curve.
cross-channel-history-retrieval
A skill for searching the original messages from earlier chat sessions and channels when the current conversation does not contain enough history.
meta-harness
Run one iteration of memory system evolution. Called by metaharness.py or interactively via /meta-harness.
llm-wiki
Build and maintain a persistent knowledge base (LLM Wiki) using native backend tools. Supports ingesting PDFs, Markdown, and TXT files, and querying or linting the data. Use this whenever the user wants to add, retrieve, or manage their own local knowledge base and documents.
memory-and-handoff
Two-mode skill: (1) session memory — load/persist durable workflow state under .cc10x/ (activeContext, patterns, progress) so context survives compaction; (2) handoff package — portable, secrets-redacted export for a coworker, different tool, or fresh non-cc10x session.