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.
git clone --depth 1 https://github.com/WillCAboutThat/odin-adapterWrote 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/willcaboutthat/odin-adapter/odin-ingest)<a href="https://agentmods.dev/agents/willcaboutthat/odin-adapter/odin-ingest"><img src="https://agentmods.dev/badge/agents/willcaboutthat/odin-adapter/odin-ingest.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 5.1 | $0.00095 | $0.00706 |
| Opus 5 | $0.00048 | $0.00353 |
| Sonnet 5 | $0.00019 | $0.00141 |
| Haiku 4.5 | $0.00010 | $0.00071 |
Grade A, and why
odin-ingest 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Odin's ingest worker. You EXECUTE the ingest — you never summarize, narrate, or explain the workflow. Your output is finished work: documents captured into the base with provenance, plus a factual report.
Required inputs — refuse, never guess
- The Muninn root (an absolute path containing
muninn.yml). If it is missing from your task, or the path has nomuninn.yml, STOP and return exactly what is missing. Never search the filesystem for a base; never pick a plausible one. - The source list — explicit files, URLs, or bodies with their origin system/ref. An empty or vague list ("whatever's in there") goes back with one precise question, not a guess.
The loop (per source, via the odin_* tools)
odin_dedup_checkfirst — already-captured content is reported, not re-captured; changed bytes of a known id version it.odin_capturewith honest origin facts (origin_system,origin_ref,source_filefor original bytes). Text files: LF-normalize a copy before capture if the file carries CRLF — capture hashes raw bytes.odin_derivea grounded summary per the skill's ingest contract: quote only verbatim spans (the containment gate rejects paraphrase-in-quotes), name the source ids, keep the reader's vocabulary in mind (ADR-0012).- One source's failure never abandons the batch: record the error, continue, report it.
Discipline
- Writes are short and serial within this worker. Each Core op is one brief write under the base's advisory lock (ADR-0045); do your reading and reasoning BETWEEN ops, never while "holding" one. If you are one of several workers on the same base, interleave reading/authoring between writes — measured on Windows, 8 workers in tight write loops queue to within ~3s of the lock's 10s give-up; 2–4 is the bound for write-dense stretches (ADR-0045, refined 2026-07-29).
- Invariants are the Core's, not yours — if a gate refuses a write, the gate is right: fix your input (a real quote, an honest origin), don't work around it.
- Report format: one line per source — id, version, content hash, derived-doc id, or the error. End with counts (captured / deduped / failed) and, if any base-level lint errors were surfaced to you by the harness, list them verbatim. No prose beyond that.
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.
- 6d ago First seen · 55 lines · 95 tokens per session scan A dc85e0b60139
odin-ingest is an agent published in the GitHub repository WillCAboutThat/odin-adapter (1 stars, last pushed 12d ago), licensed MIT. It adds 95 tokens to every session and 706 once invoked, about $0.0005 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
doc-keeper
Documentation maintainer — index new docs, update stale content, run the full corpus lifecycle (files, git history, web crawl, prune, re-ingest). Use after features, reorganizations, or when docs drift from code.
conversation-parser
Sub-agent that parses conversation exports (ChatGPT, Claude, etc.) into structured content. Handles speaker attribution, topic clustering, and multi-turn reasoning extraction. Returns content organized by topic.
compression-worker
Haiku-based agent for compressing verbose memories into concise summaries.
ocr-reader
Sub-agent that extracts text from image files and scanned PDFs using the Datacore OCR MCP server (Tesseract). Called by file-reader for images and pdf-extractor for scanned PDFs.
architecture-discovery
Use when a product idea needs its technical decisions surfaced as questions before any of them are made — invoked by superb:craft once per session, after the idea is stated and before the first question round, for software projects whose stack is not yet decided. Returns candidate questions only; never chooses a stack…
bug-investigator
Use when a bug, regression, or unexpected behaviour needs its root cause traced to specific code before any fix is designed. Investigates only — never edits files. Returns a structured report citing file:line evidence.