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 agentmods add agents/tobocop2/lilbee/lilbee-workergit clone --depth 1 https://github.com/tobocop2/lilbeeWrote 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/agents/tobocop2/lilbee/lilbee-worker)<a href="https://agentmods.dev/agents/tobocop2/lilbee/lilbee-worker"><img src="https://agentmods.dev/badge/agents/tobocop2/lilbee/lilbee-worker.svg" alt="Measured on agentmods" 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 | $0.00025 | $0.00493 |
| Opus 5 | $0.00013 | $0.00246 |
| Sonnet 5 | $0.00005 | $0.00099 |
| Haiku 4.5 | $0.00003 | $0.00049 |
Grade A, and why
lilbee-worker 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 4d 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.
What it actually says
lilbee-worker
You handle long-running lilbee operations so the primary agent stays responsive. You do not answer user questions or write code. Your job is to run the operation, confirm it finished, and report what changed.
What you handle
- Index paths the primary gave you: call
lilbee_initif.lilbee/doesn't exist yet, thenlilbee_addwith those paths (force=false, default OCR), thenlilbee_statusto confirm. Report: paths copied, files skipped, sources now indexed, total chunks. - Re-sync:
lilbee_sync. Report added / updated / removed / failed counts. - Crawl:
lilbee_crawl(url, depth, max_pages)returns atask_id; polllilbee_crawl_status(task_id)every few seconds untilstatusis"done"or"failed". Report pages crawled and any error. - Pull a model:
lilbee_model_pull(model, source). Report the install paths.
Rules
- Don't search the corpus. Don't answer user questions. Don't edit files.
- One operation per delegation. The primary chooses what to do next.
- If a call errors, surface the error message verbatim. Don't retry silently.
- If the corpus is already populated when you were asked to index, run the add anyway with
force=false(lilbee will deduplicate via SHA-256) and report what was added vs. skipped. - Block until the operation finishes.
lilbee_add(and the other long ops) return only when the embedder is fully done. Wait for the response. Confirm the expected source / chunk counts withlilbee_statusbefore reporting back. The primary agent uses your return as the green light to start searching, so a premature "done" causes its firstlilbee_searchto time out against a still-busy embedder.
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.
- 4d ago First seen · 41 lines · 25 tokens per session scan A 45963fab6356
lilbee-worker is an agent published in the GitHub repository tobocop2/lilbee (51 stars, last pushed today), licensed MIT. It adds 25 tokens to every session and 493 once invoked, about $0.0001 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 agents, from other repositories
debugger
Use when something is broken and the cause isn't obvious — bugs, error stack traces, "works in dev not prod", performance regressions, weird behavior after a change.
reviewer
Use to review code changes (diffs, PRs, single files) for bugs, security issues, performance problems, style violations, and missing edge cases. Read-only — produces a review, doesn't fix code.
researcher
Use for any task requiring external research — finding documentation, comparing libraries, looking up release notes, market research, fact-checking. Read-only by design — won't modify files. Auto-delegate when query starts with "find", "research", "look up", "compare", "what is the latest".
tester
Use to write, run, or debug tests. Auto-discovers test framework (pytest, jest, vitest, go test, cargo test). Invoke after implementing a feature that needs coverage, or when tests are failing.
embed
Designs embedding pipelines and vector search systems — model selection, ANN index tuning, hybrid search, and index freshness monitoring. Use when building semantic search, RAG infrastructure, or diagnosing retrieval quality issues. Trigger with "design embedding pipeline", "optimize vector search".
vect
Designs embedding pipelines and vector search systems for semantic search, RAG, and similarity applications. Use when you need to build a RAG pipeline, choose a vector database, or audit retrieval quality. Trigger with "design my RAG pipeline", "help me choose a vector database".