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 commands/zib-iol/the-agentic-researcher/setup_research_plangit clone --depth 1 https://github.com/ZIB-IOL/The-Agentic-ResearcherWhat 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 | $0.00007 | $0.01429 |
| Opus 5 | $0.00003 | $0.00714 |
| Sonnet 5 | $0.00001 | $0.00286 |
| Haiku 4.5 | $0.00001 | $0.00143 |
Grade A, and why
setup_research_plan 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 yesterday.
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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a research agent. This command sets up and executes a research project.
Detect which instruction file exists in the workspace and use it throughout:
- Check for:
CLAUDE.md,GEMINI.md,AGENTS.md(in that order) - Use the first one found as
$INSTRUCTION_FILE - If none exists, check if the template is available at
/claude-home/.claude/INSTRUCTIONS.md.templateand copy it toCLAUDE.md
Check the state of /workspace/$INSTRUCTION_FILE to determine what to do.
Detection logic:
- If
/workspace/$INSTRUCTION_FILEexists AND its "## 8. Project Instructions" section contains filled-in values (not just placeholders like[Research objective]), treat as RESUME. - If
/workspace/$INSTRUCTION_FILEexists but Project Instructions still has placeholders, treat as FRESH START (skip to interactive setup below). - If no instruction file exists, copy the template first, then treat as FRESH START.
Backward compatibility: If /workspace/research_instructions.md exists (from an older session), read it and migrate its contents into the Project Instructions section. Then proceed as RESUME.
RESUME (Project Instructions filled AND report.tex exists):
This is a resuming session. The project is already in progress.
- Read
/workspace/$INSTRUCTION_FILE(especially Section 8),report.tex, andTODO.md - Run
git log --oneline -20to see recent experiment commits - Run
git statusto check for uncommitted changes - Summarize the current state to the user:
- Best result so far and which experiment achieved it
- What was tried last and whether it worked
- What's next (from TODO.md or the last experiment's "Next steps")
- Ask the user if they want to continue the planned direction or pivot
- Continue the autonomous experiment loop
FRESH START (Project Instructions has placeholders, report.tex does not exist):
- Read
/workspace/$INSTRUCTION_FILESection 8 to confirm it needs filling - If report.tex exists but Section 8 is empty, read report.tex to recover context, then ask user to confirm project instructions before continuing.
- Otherwise, proceed with interactive setup below.
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.
- yesterday First seen · 123 lines · 7 tokens per session scan A 8d525bc754e6
setup_research_plan is a command published in the GitHub repository ZIB-IOL/The-Agentic-Researcher (91 stars, last pushed 2mo ago), licensed MIT. It adds 7 tokens to every session and 1,429 once invoked, about $0.0000 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 commands, from other repositories
om-vault-audit
Deep structural audit of the vault. Checks indexes, folder placement, frontmatter, links, Bases, and consistency. Fix what can be fixed, flag what needs user input.
om-incident-capture
Capture an incident from Slack channels, DMs, and threads into structured vault notes. Produces a complete incident work note with timeline, people, analysis, and brag doc entry.
om-intake
Process all unread meeting notes in work/meetings/ — reads each file, classifies content, routes to the right vault notes, then clears the inbox.
om-wrap-up
Full session review before ending. Review context, ways of working, files modified, consistency, and suggest improvements.
om-peer-scan
Deep scan a peer's GitHub PRs for performance review preparation. Produces a structured analysis saved to perf/evidence/.
om-slack-scan
Deep scan Slack channels and DMs for evidence involving a specific person or project. Extracts every touchpoint with timestamps for review evidence or project documentation.