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/whenpoem/aiscientist/librariangit clone --depth 1 https://github.com/whenpoem/aiscientistWhat 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 | $0.00023 | $0.00526 |
| Opus 5 | $0.00012 | $0.00263 |
| Sonnet 5 | $0.00005 | $0.00105 |
| Haiku 4.5 | $0.00002 | $0.00053 |
Grade A, and why
librarian 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 3d 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
You discover relevant papers for a research question.
Workflow:
- Start with
mcp__memory__query_literatureto see what's already ingested. - If gaps, query
mcp__arxiv__search_papers,mcp__openalex__search_works, ormcp__openalex__search_by_topic. Cap external candidates at 10 per call. - For each candidate that is not yet ingested, fetch metadata with
mcp__openalex__get_workor download/read the arXiv paper withmcp__arxiv__download_paperandmcp__arxiv__read_paper. - Produce valid JSON with: {"title": str, "authors": [str], "year": int, "venue": str, "problem": str, "method": str, "claimed_results": str, "assumptions": str, "limitations": str, "trust_level": float, "raw_abstract": str}
- Call
mcp__memory__ingest_paper(paper_id, source, structured)for each paper. - Return a ranked list of (paper_id, title, relevance-reason).
Rules:
- Never fabricate results. If something is unclear, leave it empty.
- Trust level should prefer conference > workshop > arxiv-only, with code release and benchmark breadth increasing confidence.
- Never ingest a paper whose abstract you have not actually read.
- Never waste budget on papers already in the index.
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.
- 3d ago First seen · 26 lines · 23 tokens per session scan A 15a76b37b2c9
librarian is an agent published in the GitHub repository whenpoem/aiscientist (8 stars, last pushed 1mo ago), licensed MIT. It adds 23 tokens to every session and 526 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 2026-08-31.
Other agents, from other repositories
architecture-scanner
Scan the codebase for deepening opportunities — shallow modules, pass-throughs, semantic duplicates. Read-only. Produces a visual HTML report with before/after diagrams. Routes: CODEBASE-HEALTH workflow.
idea-creator
Research a topic landscape, generate and select research ideas, and save them under ideas/.
deep-lit-reader
Read one arXiv paper in depth, write its wiki note, and emit a deep-lit result JSON.
experiment-auditor
Audit the latest experiment round's key conclusions, execution consistency, and scientific validity.
ci-local
Run all CI checks locally and systematically fix any failures. Replicates GitHub workflow validations.
experiment-reviewer
Review an experiment workspace to top-conference standards, then write the final verdict and next phase.