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 steerlabs/opensteer --skill knowledgegit clone --depth 1 https://github.com/steerlabs/opensteerWrote 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/steerlabs/opensteer/knowledge)<a href="https://agentmods.dev/skills/steerlabs/opensteer/knowledge"><img src="https://agentmods.dev/badge/skills/steerlabs/opensteer/knowledge/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/steerlabs/opensteer/knowledge"><img src="https://agentmods.dev/badge/skills/steerlabs/opensteer/knowledge.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00034 | $0.00654 |
| Opus 5.5 | $0.00014 | $0.00262 |
| Sonnet 5.5 | $0.00007 | $0.00131 |
| Haiku 4.5 | $0.00003 | $0.00065 |
Grade A, and why
knowledge 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 — 43 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Knowledge
Knowledge is a log of facts: one JSONB value per topic, thing and property. Each write appends a version; nothing is updated or deleted. Call sql without query first to see the columns and the readable topic ids.
facts: the current version of each fact. Use it for "what is".fact_history: every version, withreplaces_idpointing at the one it replaced. Use it for "what was", "who changed" and "when".textisvalueas plain text ("active"becomesactive). Compare and cast it:text = 'active',text::numeric > 2000. Nevervalue::text, which keeps JSON quotes and silently matches nothing.- Always filter by
topic, and join two properties of one thing ontopicandthing. The same thing key in another topic is a different thing. - Select
writer_kind,writer_idandwritten_atwhen you say who recorded something. An aggregate has no single writer. - Parameters are JSON. Send a list as a JSON array and read it as JSON:
thing IN (SELECT jsonb_array_elements_text($2::jsonb)).= ANY($2)and$2::text[]are refused.
SELECT text, writer_id, written_at FROM facts
WHERE topic = $1 AND thing = 'trip/kyoto' AND property = 'hotel';
SELECT s.thing, b.text::numeric AS budget FROM facts s
JOIN facts b ON b.topic = s.topic AND b.thing = s.thing AND b.property = 'budget'
WHERE s.topic = $1 AND s.property = 'status' AND s.text = 'active';
SELECT thing FROM facts
WHERE topic = $1 AND property = 'profile' AND value @> '{"country": "US"}';
SELECT id, text, replaces_id, writer_id, written_at FROM fact_history
WHERE topic = $1 AND thing = 'trip/kyoto' AND property = 'hotel' ORDER BY id;
A refusal names what isn't allowed: rewrite with plain SELECT, joins, aggregates, window functions and common text, date and JSON functions. Catalogs, settings and other tables are not there to find.
Writing
Before an INSERT, read writing: replacing a fact, onlyIfUnchanged, receipts and uncertain writes.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 43 lines · 34 tokens per session scan A 7cbfdce7045c
knowledge is a skill published in the GitHub repository steerlabs/opensteer (213 stars, last pushed today), licensed MIT. It adds 34 tokens to every session and 654 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-11-11.
knowledge compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| knowledgesteerlabs/opensteer | Skill | 213 | today | MIT | A | 34 | Codex |
| onboarding-milestoneskortix-ai/suna | Skill | 21k | 4d ago | none | A | 62 | Claude Code, Codex |
| dsar-fulfillmentkortix-ai/suna | Skill | 21k | 4d ago | none | A | 85 | Claude Code, Codex |
| weekly-metrics-reportkortix-ai/suna | Skill | 21k | 4d ago | none | A | 47 | Claude Code, Codex |
| setuptimurgaleev/memex | Skill | 11 | 1mo ago | MIT | A | 25 | Claude Code, Codex |
| butterbasebutterbase-ai/butterbase | Skill | 3.0k | 3d ago | Apache-2.0 | A | 34 | Claude Code, Codex |
| build-with-tinybasetinyplex/tinybase | Skill | 5.2k | 21d ago | MIT | A | 90 | Claude Code, Codex |