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/ourobx/agent-ebpf/claude-mdgit clone --depth 1 https://github.com/ourobx/agent-ebpfWrote 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/instructions/ourobx/agent-ebpf/claude-md)<a href="https://agentmods.dev/instructions/ourobx/agent-ebpf/claude-md"><img src="https://agentmods.dev/badge/instructions/ourobx/agent-ebpf/claude-md.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 | $0.01424 | $0.01424 |
| Opus 5 | $0.00712 | $0.00712 |
| Sonnet 5 | $0.00285 | $0.00285 |
| Haiku 4.5 | $0.00142 | $0.00142 |
Grade A, and why
agent-ebpf CLAUDE.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 3d 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.
This is a copy
86% identical to agent-ebpf copilot-instructions.md — 42 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 20 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Installed Claude Skills
Claude Code discovers and loads skills under .claude/skills/ automatically — nothing here needs to be read for that to work. This table is kept up to date purely as a human-readable summary of what's installed and why.
| Skill | Detected via | Description |
|---|---|---|
| cross-platform-scripting | **/*.ps1, **/*.psm1, **/*.sh, **/*.cmd, **/*.bat |
Detect the host OS (Windows/macOS/Linux) and PowerShell version (5.1 Desktop vs 7+ Core) before writing or editing scripts, and write/adapt .ps1, .sh, and .cmd scripts to match what's actually available — avoiding PS7-only syntax on PS5.1, GNU-only flags on macOS/BSD tools, and Windows-only assumptions on POSIX shells. Use before writing a new script, when a script fails with a syntax/parameter error that looks version- or OS-specific, or when asked to make a script cross-platform. |
| cursor-kiro-extension-publishing | **/PUBLISHING.md, **/publish-openvsx.js, **/.github/workflows/publish*.yml, **/00-extension-registries.md |
Publish a VS Code-compatible extension to Open VSX for Cursor and Kiro IDE using ovsx. Covers same VSIX as VS Marketplace, OVSX_PAT setup, namespace ownership verification, publish scripts, GitHub Actions, wrong-VSIX pitfalls, and cross-links between registries. Use when publishing to Open VSX, Cursor gallery, Kiro extension registry, or debugging ovsx publish failures. Pair with vscode-extension-publishing for packaging and VS Marketplace. |
| deployment-practical | **/*.tf, **/*.bicep, **/azure.yaml, **/azure.yml, **/Dockerfile, **/Dockerfile.*, **/docker-compose*.yml, **/.gitlab-ci.yml, **/azure-pipelines.yml, **/.env*, **/deployment/** |
Deployment-first delivery — concrete architecture and IaC over theoretical advice. Use when deploying, provisioning infra, debugging first-apply failures, or when the user wants advice that works on the first attempt (not hand-wavy theory). Pair with Practical Focus toggle (architecture-first / deploy-ready). |
| doc-coauthoring | **/*.md, **/docs/** |
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. |
| extension-value-audit | **/.claude/learning/hook-health.jsonl, **/.claude/learning/task-skill-proposals.json, **/.claude/learning/skill-adoption.jsonl, **/.claude/learning/runs.jsonl |
Give an honest, evidence-based verdict on whether the "Claude Skills Manager" VS Code extension (claude-skill-deployer) is actually delivering real value in THIS project — not a usage-count report of individual skills (see skill-usage-insights for that), but a judgment of the extension's own core mechanisms (adaptive skill suggestions/confidence scoring, task-focus, budget/cost gating, MCP-Force security mode, cross-agent sync, hooks pipeline) against real .claude/learning/ telemetry and the agent's own actual session experience. Use when asked "is this extension helping", "give feedback on Claude Skills Manager", "extension ROI", "does the extension actually do anything", "audit the extension's value", or similar. Produces a blunt, numbers-grounded report — not marketing copy. |
| file-style-conventions | **/* |
Apply two lightweight file-hygiene conventions when writing or editing files - no emoji characters outside Markdown (.md) files, and YAML files (.yml/.yaml) end with exactly one trailing newline. Use whenever creating or editing non-Markdown files that might contain emoji, or any .yml/.yaml file. |
| github-actions-ci | **/.github/workflows/*.yml, **/.github/workflows/*.yaml, **/.github/**/*.md, **/test/**, **/src/**, **/*.test.ts, **/*.test.js |
Debug GitHub Actions pipeline failures and reproduce CI stages locally. Use when asked to debug CI, fix a failing workflow, reproduce a job with act, or run a pre-flight check before pushing. |
| self-learning | **/* |
Maintain a project-local self-learning base of task/command outcomes — record successes and failures with timestamps, durations, and fixes; generate a patterns report (pass rates, recurring errors, known fixes); and surface a learned hint before retrying something that failed before. Use at the start of a session to check learned hints, after running a non-trivial command/skill to record the outcome, when asked "what failed before" or "what did we learn", or to record a manual decision/learning. |
| skill-feedback-adaptation | **/.claude/learning/skill-feedback.jsonl, **/.claude/learning/task-skill-proposals.json, **/.claude/learning/** |
AUTO-START on new agent session/window (injected by profile-init-watch for Claude, Cursor, Kiro, Copilot) and on new tasks — analyze the prompt and repo, write task-skill-proposals.json, then read top proposed skills before other work. Also register user disagreement into skill-feedback.jsonl when the user says no, not, wrong, stop, or disagrees with agent output. |
| skill-official-updater | **/* |
At the start of a new session, do a cheap check for new or updated official Anthropic skills (github.com/anthropics/skills) and automatically add or update them in skills_library/ (no user prompt). Also use on explicit request ("check for official skill updates", "sync official skills"). |
| skill-usage-insights | **/.claude/learning/runs.jsonl, **/.claude/skills/** |
Analyze recorded skill usage in this project (.claude/learning/runs.jsonl, written by self-learning) and the skills installed in .claude/skills/ to produce a usage and KPI report - which skills are actively used and reliable, which are failing, and which are unused or low-value, with recommendations on what to add or remove. Use when asked for "skill usage stats", "skill KPIs", "which skills should we add or remove", or "are our installed skills still useful". |
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.
- 3d ago First seen · 20 lines · 1,424 tokens per session scan A 1778de43437c
agent-ebpf CLAUDE.md is an instructions file published in the GitHub repository ourobx/agent-ebpf (1 stars, last pushed 8d ago), licensed MIT. It adds 1,424 tokens to every session, about $0.0071 per session on Opus 5. A static security scan graded it A with 0 findings. It is 86% identical to agent-ebpf copilot-instructions.md, differing in 42 lines, and is treated as a copy.
Other instructions, from other repositories
casbin-gateway CLAUDE.md
Claude Code instructions for apache/casbin-gateway, covering claude.md and code style.
aigis CLAUDE.md
Instructions for killertcell428/aigis, covering aigis — project guidelines, zenn記事管理（zenn cli + github連携）, ディレクトリ構成, 記事の作成 and プレビュー.
preloop AGENTS.md
AGENTS.md instructions for preloop/preloop, covering preloop development guide, commands, git workflow, keep sample data generic and code style.
pfsense-mcp-server AGENTS.md
Instructions for night4me/pfsense-mcp-server, covering pfsense mcp server, architecture and security, development workflow, long-running validation and test parallelism.
xops-cli AGENTS.md
Instructions for wentf9/xops-cli, covering ai 助手开发协作指南 (agents.md), 🎨 1. 编码规范 (coding standards), 📦 2. git 规范 (git standards) and 🛠️ 3. 测试、构建与 ci (testing, building & ci).
fwskillsshare AGENTS.md
Instructions for fastrevmd-lab/fwskillsshare, covering firewall skills repository instructions, purpose and architecture, setup and development, required checks and skill description constraints.