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 bgauryy/octocode --skill octocode-documentationgit clone --depth 1 https://github.com/bgauryy/octocodeWrote 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/bgauryy/octocode/octocode-documentation)<a href="https://agentmods.dev/skills/bgauryy/octocode/octocode-documentation"><img src="https://agentmods.dev/badge/skills/bgauryy/octocode/octocode-documentation/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/bgauryy/octocode/octocode-documentation"><img src="https://agentmods.dev/badge/skills/bgauryy/octocode/octocode-documentation.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 | $0.00107 | $0.01966 |
| Opus 5 | $0.00053 | $0.00983 |
| Sonnet 5 | $0.00021 | $0.00393 |
| Haiku 4.5 | $0.00011 | $0.00197 |
Grade C, and why
octocode-documentation scanned grade C with 1 finding 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 12d 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
Suppression: `<!-- style-lint: ignore-file -->` skips a file found by recursion, `<!-- style-lint: ignore-line rule-id -->` mutes named rules on one line. Both are inert inside a code span, so a page can document them. D How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Octocode Documentation
Evidence-backed docs for humans and agents, written to the Google developer documentation style guide. Classify first. Gate writes. Prefer durable cross-refs over code dumps.
Flow
UNDERSTAND → RESEARCH → CLASSIFY → OUTLINE GATE → WRITE → STYLE → VERIFY
UNDERSTAND names the deliverable, audience, and target paths. Compress when the request already names targets and type. Expand when claims need verification. A copyedit request starts at STYLE. Answer a single-term question ("is allows you to okay?") straight from assets/google-word-list.tsv — quote the guidance and stop.
Rules
- Verify claims in the repository before asserting them. Invented commands, paths, APIs, and env vars are the one unrecoverable failure — omit or mark "Not verified in repository" instead.
- Pick one mode and load its routes before writing.
- Gate creates and overwrites unless the requester approved the targets this turn; a copyedit of a named file carries its own approval. Touch only the files they named — propose the rest.
- Apply the style defaults in
references/style-index.mdto every line you write or edit, and name the rule when you change someone else's wording. - The style references are a snapshot of Google's guide, and every one links the pages it restates. IF someone disputes a rule, the pack doesn't carry it, or the wording carries risk (trademark, product name, legal claim, security claim) → THEN open the live page with a web tool, quote it with its URL, and fix the reference when it disagrees; say so when no fetch was possible.
- A style pass changes wording, not claims; a fact change goes back to RESEARCH.
AGENTS.mdis an index of links and non-obvious rules, not a content dump.- Prefer durable pointers (module path, contract name, doc link) over line numbers and pasted code.
- One Diátaxis type per page; link siblings instead of mixing.
- IF the project documents its own style guide, or the repository already applies a convention consistently → THEN follow it and report the conflict instead of adding a second scheme.
What ships with it
35 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.
- assets/google-style-pages.tsv 6.4 KB
- assets/google-word-list.tsv 104 KB
- README.md 4.9 KB
- references/adr.md 1.5 KB
- references/agent-readable.md 1.7 KB
- references/agents-md.md 1.8 KB
- references/diataxis.md 1.4 KB
- references/evidence-research.md 1.6 KB
- references/modes.md 2.2 KB
- references/style-abbreviations.md 2.7 KB
- references/style-api.md 3.1 KB
- references/style-blocks.md 3.1 KB
- references/style-claims.md 3.8 KB
- references/style-cli.md 3.7 KB
- references/style-code.md 3.2 KB
- references/style-examples.md 2.5 KB
- references/style-format.md 4.3 KB
- references/style-global.md 3.4 KB
- references/style-grammar.md 4.1 KB
- references/style-images.md 3.1 KB
- references/style-inclusive.md 3.3 KB
- references/style-index.md 4.0 KB
- references/style-links.md 3.3 KB
- references/style-numbers.md 5.1 KB
- references/style-procedures.md 2.6 KB
- references/style-punctuation.md 5.8 KB
- references/style-review.md 2.8 KB
- references/style-sources.md 3.3 KB
- references/style-structure.md 3.9 KB
- references/style-ui.md 3.8 KB
- references/style-voice.md 4.7 KB
- references/style-words.md 3.1 KB
- references/write-verify.md 2.3 KB
- scripts/refresh-word-list.mjs 7.2 KB runs code
- scripts/style-lint.mjs 24 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.
- 12d ago First seen · 84 lines · 107 tokens per session scan C 4ae90d3fccdf
octocode-documentation is a skill published in the GitHub repository bgauryy/octocode (932 stars, last pushed today), licensed MIT. It adds 107 tokens to every session and 1,966 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
create-handler
Create a new API handler following the handler pattern. Use when the user asks to create an endpoint, handler, or route.
architecture-docs
Workflow for creating and maintaining architecture documentation. Use when the user needs to document system architecture or make ADRs.
component-creation
Step-by-step workflow for creating accessible, tested UI components. Use when the user asks to create a new UI component.
fix-tests
Systematic workflow for diagnosing and fixing failing tests. Use when the user reports failing tests or asks to fix test failures.
infrastructure
Workflow for provisioning cloud infrastructure using IaC. Use when the user needs to create or modify cloud resources.
migration
Safe workflow for creating and applying database schema changes. Use when the user needs to modify the database schema.