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 instructions/robcsaszar/delve/agents-mdgit clone --depth 1 https://github.com/robcsaszar/delveWhat 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 | $0.00158 | $0.00158 |
| Opus 5 | $0.00079 | $0.00079 |
| Sonnet 5 | $0.00032 | $0.00032 |
| Haiku 4.5 | $0.00016 | $0.00016 |
Grade A, and why
delve AGENTS.md 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.
What it actually says
AGENTS.md
Mission
This repo publishes a single skill, delve, copied in from a source repo (Orakl) where it is used and refined, then published here for anyone to install. There is no build, no tests, no runtime of its own. The deliverable is skills/delve/.
Judgment boundary
Don't let the skill's content drift toward being useful only in one narrow context. It exists to make any agent a more careful, confidence-flagging research collaborator on any topic; a change that only makes sense for one specific domain or project doesn't belong here. If you're updating it, re-run a grep for source-repo-specific leakage (grep -rni "orakl" under skills/) before publishing.
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 · 10 lines · 158 tokens per session scan A 664e32343d20
delve AGENTS.md is an instructions file published in the GitHub repository robcsaszar/delve (0 stars, last pushed 2d ago), licensed MIT. It adds 158 tokens to every session, about $0.0008 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 instructions, from other repositories
FirstData AGENTS.md
Instructions for MLT-OSS/FirstData, covering agents.md, what this repo is, validation, install dependencies (first time only) and run all validation checks.
bullshit-detector CLAUDE.md
Instructions for SerhiiKorniienko/bullshit-detector: Skills are organized into bucket folders under skills/.
contentforge AGENTS.md
Instructions for indranilbanerjee/contentforge, covering contentforge — agent context, what this plugin is, how to use it as an agent, canonical entry points and files in this repo.
statcite CLAUDE.md
Claude Code instructions for asokore/statcite, covering claude.md: statcite project guide, commands (run in server/), architecture (server/src) and invariants: do not break.
fable5-codex AGENTS.md
Instructions for rhein1/fable5-codex, covering fable-5 repo instructions, working method and package rules.
statcite GEMINI.md
Gemini CLI instructions for asokore/statcite, covering statcite: verify economic statistics before you state them, the rule, writing with economic numbers, fact-checking a draft and when sources disagree.