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 fvadicamo/dev-agent-skills --skill privacy-guardgit clone --depth 1 https://github.com/fvadicamo/dev-agent-skillsWrote 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/fvadicamo/dev-agent-skills/privacy-guard)<a href="https://agentmods.dev/skills/fvadicamo/dev-agent-skills/privacy-guard"><img src="https://agentmods.dev/badge/skills/fvadicamo/dev-agent-skills/privacy-guard.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00128 | $0.02478 |
| Opus 5 | $0.00064 | $0.01239 |
| Sonnet 5 | $0.00026 | $0.00496 |
| Haiku 4.5 | $0.00013 | $0.00248 |
Grade A, and why
privacy-guard 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 8d 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 — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
privacy-guard
Guard against accidentally publishing details of your private infrastructure in public repositories (or in repositories that may become public later). Two legs: behavioral rules for the Claude session, always on, and a per-repo technical gate: a pre-commit hook backed by a gitignored local denylist, plus gitleaks.
Threat model
The public repository is the boundary. Anything that describes your private infrastructure must not cross it, in any form: committed files, commit messages, branch names, PR and issue bodies, release notes, published artifacts, screenshots.
Sensitive, never in a public repo:
- node hostnames and internal aliases (workstations, VPS, client machines)
- names of internal projects and service instances that were never published
- local usernames, personal and work email addresses
- personal domains and URLs of internal services
- absolute paths of home directories and internal mounts
- private-network IP ranges and VPN address space (for example the Tailscale CGNAT range
100.64.0.0/10) - cookie files, tokens, credentials (also covered by gitleaks)
Not sensitive, fine in public docs:
- public products and technologies named generically (Tailscale, Docker, n8n)
- the maintainer's public GitHub username
- deployment patterns described in generic form ("a Docker host reachable over your private network / behind a VPN")
The concrete list of sensitive tokens is yours and stays private. This skill ships
references/denylist-template.txt, a placeholder-only starting point: fill it in a private
location (a private dotfiles repo, a private notes repo) and treat that filled copy as the
seed you propagate from. Never commit the filled version to a public repo.
Behavioral rules (Claude session in a public repo)
- Never write internal tokens into committed files. Private deployment details are
documented in a private repo or in a gitignored
.local/directory, never in the public repo's docs. - The user mentioning internal names in conversation does NOT authorize writing them into the repo: the conversation is private, the repo is not.
- Before every commit: re-read the diff looking for denylist tokens. The hook is the safety net, not the first check.
- This applies outside files too: commit messages, branch names, PR and issue titles and bodies, CHANGELOG entries, published artifacts.
- If a sensitive token has already reached published git history: tell the user immediately (rotation, history rewrite with BFG or filter-repo are their calls), do not just remove it from the tip.
What ships with it
4 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.
- 8d ago First seen · 180 lines · 128 tokens per session scan A 63990d6f0dd3
privacy-guard is a skill published in the GitHub repository fvadicamo/dev-agent-skills (72 stars, last pushed 2d ago), licensed MIT. It adds 128 tokens to every session and 2,478 once invoked, about $0.0006 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 2026-08-30.
Other skills, from other repositories
bulwark-brainstorm
Role-based brainstorming with dual modes: --scoped (sequential Task tool, 5 roles) and --exploratory (Agent Teams peer debate, 4 roles). Use for feasibility assessment and idea validation.
plan-creation
Create structured implementation plans via a 4-role scrum team (Product Owner, Architect, Eng/Delivery Lead, QA/Critic) with optional Agent Teams peer debate mode.
anthropic-validator
Validates Claude Code assets (skills, hooks, agents, commands, MCP servers, plugins) against official Anthropic standards. Fetches latest docs dynamically and produces structured validation reports.
code-review
Comprehensive code review with distinct aspect based sections. Use when reviewing code, checking for security issues, finding type safety problems, auditing code quality, or when user asks to review code, PRs or changes. Three-phase workflow runs static tools, LLM judgment, and writes diagnostic log.
create-subagent
Generates single-purpose Claude Code sub-agents for use via the Task tool. Use when creating dedicated sub-agents, scaffolding agent definitions, or generating agents with diagnostics and permissions setup.
test-audit
Audit test suites for T1-T4 violations using AST analysis, mock detection, and multi-stage synthesis. Invoke when user asks to audit tests, check test quality, find mock violations, review test effectiveness, or inspect test suites for over-mocking. Triggers automatic rewrites when quality gates fail.