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 bks-lab/open-bridge --skill secretsgit clone --depth 1 https://github.com/bks-lab/open-bridgeWrote 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/bks-lab/open-bridge/secrets)<a href="https://agentmods.dev/skills/bks-lab/open-bridge/secrets"><img src="https://agentmods.dev/badge/skills/bks-lab/open-bridge/secrets/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/bks-lab/open-bridge/secrets"><img src="https://agentmods.dev/badge/skills/bks-lab/open-bridge/secrets.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 | $0.00076 | $0.03488 |
| Opus 5.5 | $0.00030 | $0.01395 |
| Sonnet 5.5 | $0.00015 | $0.00698 |
| Haiku 4.5 | $0.00008 | $0.00349 |
Grade A, and why
secrets 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 19d 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 — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Secrets
Owns every secret reference written down in this tree: what they are, whether they still resolve here, and how a value reaches a program without reaching the answer. Read the referenced file ONLY when triggered.
The engine is skills/secrets/engine/ behind the shim
skills/secrets/secrets.sh. The grammar it parses is the machine readable twin
of rules/secret-placement.md, which stays
the source of truth for which schemes exist and where a secret belongs.
The principle
An agent reads this output. Whatever is printed here sits in the model's context for the rest of the session, in the transcript, and in whatever log the harness keeps. So this skill hands out three things about a secret and never a fourth: its NAME (the reference), its LENGTH in bytes, and a FINGERPRINT (the first eight hex characters of its sha256). That is enough to tell two live tokens apart, to see that a rotation actually landed, and to prove that something was read. It is not enough to use.
There is deliberately no command that prints a value. Not behind a flag,
not behind a confirmation, not in --json. run is the one path a value
takes, and it goes into a child process rather than into the answer: the
child's own output is scrubbed on the way back, and a stream that still holds
the value after scrubbing is dropped rather than printed.
The second thing the report carries is the one nobody had: CONTEXT. The same
reference is readable from a desktop session and refused over ssh. A report
that says ok on a laptop and missing on the same laptop over ssh is not
measuring the vault, it is measuring the session, and it has to say which.
Where a new secret goes
The other half of the question, and the half that had no answer. An agent handed
a token decided for itself where to keep it, and the result was small text files
with credentials in working folders on two machines. The answer is declared
once, in infra/secret-stores/*.yaml: one file per store, saying where that
store is, who reaches it from where, and which KIND of secret belongs in it.
What ships with it
52 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.
- .gitignore 34 B
- engine/__init__.py 361 B runs code
- engine/audit.py 9.6 KB runs code
- engine/backends/__init__.py 2.0 KB runs code
- engine/backends/azure_keyvault.py 6.6 KB runs code
- engine/backends/base.py 5.2 KB runs code
- engine/backends/file.py 8.0 KB runs code
- engine/backends/keepass.py 11 KB runs code
- engine/backends/keychain.py 12 KB runs code
- engine/backends/onepassword.py 4.5 KB runs code
- engine/check.py 7.4 KB runs code
- engine/cli.py 26 KB runs code
- engine/discover.py 6.8 KB runs code
- engine/errors.py 3.2 KB runs code
- engine/exec.py 5.5 KB runs code
- engine/patterns.py 9.3 KB runs code
- engine/refs.py 13 KB runs code
- engine/resolve.py 11 KB runs code
- engine/stores.py 14 KB runs code
- engine/values.py 8.8 KB runs code
- references/audit.md 20 KB
- references/resolve.md 22 KB
- references/store.md 15 KB
- run-tests.sh 212 B runs code
- scripts/tests/mutate.py 4.7 KB runs code
- scripts/tests/run.sh 2.3 KB runs code
- scripts/tests/tally.awk 3.1 KB
- secrets.sh 2.7 KB runs code
- tests/__init__.py 611 B runs code
- tests/conftest.py 25 KB runs code
- tests/mutations.py 86 KB runs code
- tests/test_acceptance.py 42 KB runs code
- tests/test_audit_cli.py 39 KB runs code
- tests/test_audit.py 47 KB runs code
- tests/test_azure_keyvault.py 34 KB runs code
- tests/test_check.py 24 KB runs code
- tests/test_cli.py 25 KB runs code
- tests/test_discover.py 24 KB runs code
- tests/test_file_backend.py 30 KB runs code
- tests/test_keepass_write.py 30 KB runs code
- tests/test_keepass.py 42 KB runs code
- tests/test_keychain_path_shape.py 7.1 KB runs code
- tests/test_keychain.py 33 KB runs code
- tests/test_onepassword.py 24 KB runs code
- tests/test_patterns.py 38 KB runs code
- tests/test_refs.py 30 KB runs code
- tests/test_resolve.py 33 KB runs code
- tests/test_store_cli.py 48 KB runs code
- tests/test_stores.py 52 KB runs code
- tests/test_values.py 28 KB runs code
- tests/test_where.py 32 KB runs code
- tests/test_write_keychain.py 34 KB runs code
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.
- 19d ago Changed · -6 lines · -113 tokens per session c0b9f6b2e1a2
- 25d ago First seen · 245 lines · 189 tokens per session scan A 4898e0506f8d
secrets is a skill published in the GitHub repository bks-lab/open-bridge (13 stars, last pushed today), licensed MIT. It adds 76 tokens to every session and 3,488 once invoked, about $0.0003 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 2026-09-21.
Other skills, from other repositories
agentmail
Use your assigned AgentMail inbox to read email tasks, explicitly send or reply, and check delivery. Provided automatically by your inbox assignment.
mcp-host-styling-integration
Integrates MCP App UI with host theming system. Applies host CSS variables, handles onhostcontextchanged, safe area insets, display mode detection, and fullscreen configuration.
dd-synthesis
Synthesize findings from multiple specialist agents into a coherent investment recommendation. Use when combining reports from 9+ specialist agents into a single decision document.
cog-knowledge-consolidation
Build structured knowledge frameworks from scattered vault notes with source attribution.
formula-authoring
Author TOML-based Formula workflow templates that become Protomolecules and active Molecules in Gas Town's durable workflow system.
incident-postmortem-report
Produce a thorough incident post-mortem report after an outage or customer-impacting event. Covers executive summary, impact, detailed timeline, root cause, contributing factors, corrective and preventive actions, and lessons learned. Use when the user asks to write, draft, or complete a post-mortem, blameless review…