agentleFS
BirdizBirdiz/skills/skills/audit/recon/SKILL.md

recon

Skill2 starsChanged 15 days agoMITView on GitHub

Map an unfamiliar codebase before auditing it (stacks, entry points, modules, execution flows, trust boundaries, personal-data flows, glossary). Use at the start of an audit, or when an audit axis needs codebase context and no codebase map exists yet.

32 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use reporting, osint and 3 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • GitHubSkill

    reporting

    Report formatting guidelines for gh-aw workflows (header levels, collapsible details, run-ID links)

    Also in 3 projects that use this

  • JosBen74Skill

    osint

    OSINT-sökning och informationsinhämtning från öppna källor för Länsstyrelsen VG

    Also in 4 projects that use this

  • samklee1Skill

    web-pentest

    Orchestrates authorized web app pentesting on Kali Linux using Opus sub-agents with difficulty-aware planning (TDA/EGATS), browser-first exploration, iterative verification, two-layer review (per-sub-agent + phase-level), optional cross-model gap audit (GPT/codex via codex-plugin-cc), and enforced KB updates. Trigger on: pentesting, penetration testing, web app security assessment, vulnerability assessment, OWASP testing, PTES, attack surface mapping, API security testing, pentest reporting, or 'help me test this app'. NOT for network-only pentesting, malware analysis, reverse engineering, or CTFs.

    Also in 3 projects that use this

  • SentrySkill

    create-pr

    Create pull requests following Sentry conventions. Use when opening PRs, writing PR descriptions, or preparing changes for review. Follows Sentry's code review guidelines.

    Also in 3 projects that use this

  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 27 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every statement carries a pointer: file and line, or a query and its result.

    line 89

“Common” is measured: kept by at least 6 in 10 of the 27 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

recon

Version: 0.26.0 (audit skills, see CHANGELOG)

Build the codebase map once, so every axis starts from the same understanding. Input: 00-engagement.md (missing: stop, ask the auditor to run /birdiz-skills:audit-kickoff). Output: <workspace>/01-codebase-map.md. Bound by rules of engagement.

Without the repository capability, map what the running target shows and say so at the top.

Working copy and dependencies

Work in the dedicated clone at the audited commit that kickoff recorded as working_copy; without one, create it and give its path in the end report. For each ecosystem of the project (Composer, npm, an import map...), take the dependencies' code from the first available:

  1. an existing install with a byte-identical lockfile (cmp) whose record of installed packages lists every locked one, path recorded;
  2. an install into the clone that runs no package code (composer install --no-scripts --no-plugins, npm ci --ignore-scripts, or equivalent), the lockfile unchanged afterwards (git status);
  3. neither: the questions left open go under Unknowns.

An ecosystem that only holds tooling outside the scope (test runners, agent tools) is recorded as such and not installed.

Strategy

With the profile's graph_tool, take structure from it (clusters, call chains, flows), then read the code at every point the map relies on: a graph can hold wrong edges, and each one goes under Unknowns. Without it, or with its server unreachable, use its CLI, else go breadth first: manifests, entry points, then inward until each module's role is clear. The map is the same either way.

The map

# Codebase map
commit: <sha>
method: <graph tool name@version | search>
dependencies:
  <ecosystem>: <path of identical install | installed without scripts | absent | out of scope>
skill: recon@<version>
tools: [<name@version>]
generated: <YYYY-MM-DD>

## Stacks
## Entry points
## Modules
## Main execution flows
## Trust boundaries
## Personal data
## External dependencies
## Glossary
## Unknowns
  • Stacks: languages, frameworks, runtimes, versions, where declared.
  • Entry points: routes, commands, consumers, jobs, webhooks; file:line each.
  • Modules: real units and duties; where structure departs from appearance.
  • Main execution flows: those that carry the product, usually 5 to 12, entry to side effect, each with the pages or routes it covers, so every axis reads "main flow" the same way.
  • Trust boundaries: untrusted input, authn and authz enforcement, outbound calls.
  • Personal data: what; where it enters, is stored, leaves (logs, third parties, exports).
  • External dependencies: services, APIs, stores, infrastructure.
  • Glossary: domain terms as the code uses them.
  • Unknowns: numbered; what, why it matters, "Settled by: <agent | auditor | owner>: <how>".

The agent settles by reading code or running an analysis tool, and by running the project when the engagement's execution allows it; otherwise running the project, and always stating an intent, falls to the auditor or owner (for an auditor-owned target, the auditor may paste a command's output).

Rules

  • Every statement carries a pointer: file and line, or a query and its result. The map's pointers need no quote; findings and leads quote theirs.
  • A graph edge the map relies on is checked; one found wrong goes under Unknowns. Edges the map does not use are not inventoried.
  • A graph tool may register its index outside the workspace (a registry entry): record where. It may also write into the copy: run GitNexus with gitnexus analyze --index-only, whose default edits AGENTS.md, CLAUDE.md and skills; for another tool read --help, and check git status afterwards.
  • The map describes. Problems seen on the way become short leads (observation, pointers, settling check) in the relevant findings/<axis>.md, in the engagement language; an axis outside the plan gets its file with coverage not-requested.
  • A call no method shows is written "no caller found by <method>", under Unknowns, like every other gap.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.