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 dgilford/ai-science-toolkit --skill pickupgit clone --depth 1 https://github.com/dgilford/ai-science-toolkitWrote 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/dgilford/ai-science-toolkit/pickup)<a href="https://agentmods.dev/skills/dgilford/ai-science-toolkit/pickup"><img src="https://agentmods.dev/badge/skills/dgilford/ai-science-toolkit/pickup.svg" alt="Measured on agentmods" 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 | $0.00065 | $0.00514 |
| Opus 5 | $0.00032 | $0.00257 |
| Sonnet 5 | $0.00013 | $0.00103 |
| Haiku 4.5 | $0.00006 | $0.00051 |
Grade A, and why
pickup 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 8d 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
Resume this project from the repo-local handoff state.
Current repo state
git status --short 2>/dev/null || echo "(not a git repo)"
git log --oneline -12 2>/dev/null || echo "(no git log)"
Instructions
Check for .ai/HANDOFF.md. Two paths:
If found: Read it, along with CLAUDE.md and AGENTS.md if present. Then verify it against current repo state: confirm the branch matches, check that referenced files exist, and flag any commits more recent than the handoff that aren't reflected in it. If .ai/reviews/ exists, note the most recent review report and surface any of its unresolved findings that bear on the next action.
If not found: Tell the user explicitly that no handoff was found before doing anything else. Then reconstruct best-effort from the repo: scan recent git log, read recently modified files, and check for obvious entry points (scripts, notebooks, configs).
Report
State on the first line whether the handoff was loaded: Handoff loaded from .ai/HANDOFF.md or No handoff found — reconstructed from repo.
Then open with a 2–3 sentence conversational recap — what this project is, what was being worked on, and where things stand. Write it the way a colleague would catch someone up after they've been away, not as a bullet list.
Then follow with only the sections that have real content:
Key decisions — consequential choices already locked in.
Scientific context — datasets, baselines, reference periods, counterfactual definitions, known data quality issues.
Risks — load-bearing assumptions, things likely to go wrong.
Next action — the one concrete thing to do right now. If multiple paths are plausible, compare them in 2–3 lines each and recommend one.
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.
- 8d ago First seen · 46 lines · 65 tokens per session scan A fae3a3957ba5
pickup is a skill published in the GitHub repository dgilford/ai-science-toolkit (62 stars, last pushed 19d ago), licensed MIT. It adds 65 tokens to every session and 514 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
alive-morning
Morning briefing -- read all walnut states, surface priorities, inbox count, stale walnuts, people nudges.
notion
Read, create, and update Notion pages and databases via the Notion REST API.
wrap-session-up
End-of-session review for a Claude Code conversation. Replays what was discussed, checks for open action items, documents relevant outcomes in the brain vault, and creates Todoist tasks for anything untracked. Use at the end of a work session to ensure nothing falls through the cracks.
shipkit-work-memory
Log session progress and save resume state. Infers from conversation and git. Triggers: 'log progress', 'session summary', 'checkpoint', 'save progress', 'end session'.
pm-setup
Interactive onboarding wizard that creates the PM-OS knowledge base for your product. Asks questions, fetches company info, and populates all knowledge files.
mxSave
Use when the user says "save state", "/mxSave", "session end", "before /compact", "wrap up", or otherwise wants to persist the current mx-project state (clean settings, update CLAUDE.md + docs/status.md, create session notes in MCP-DB, sync orchestrate-state deltas, emit clear-cycle tip). Loop-capable. Fires at…