Recon-Skills

What it is

Recon-Skills (github.com/uphiago/recon-skills, MIT) is a repository of markdown SKILL.md playbooks for authorized external security testing, maintained by the author of hiago.sh. Beyond the category layout the summary gives, the page specifies a concrete adoption workflow, recommended entry-point skills, an output convention, and a validator — the operational details you need before cloning.

How it works

Skills are plain files you can follow by hand or load as task context into an automation/AI system. Each is discoverable by filename (SKILL.md) and searchable by technique keyword. The hunt-* skills each cover one vulnerability class or platform surface. The intended usage is discovery-first: run broad recon skills, then add a vulnerability or platform skill only once discovery produces a matching signal — not by loading everything up front.

What to do

Clone and enumerate the available skills:

git clone https://github.com/uphiago/recon-skills.git
cd recon-skills
find . -name SKILL.md -print | sort
rg -n "SSRF|OAuth|GraphQL|Kubernetes" --glob 'SKILL.md'

For a broad external web assessment, start from these entry points: redteam/web2-recon, recon/subdomain-enumeration, recon/web-enumeration, and redteam/bb-methodology.

Skills write persistent artifacts under ${OUTPUT_DIR:-./output}, so set a writable location first:

export OUTPUT_DIR="${OUTPUT_DIR:-./output}"
mkdir -p "$OUTPUT_DIR"

If contributing or reviewing a change, run the catalog validator — structural errors fail the command; style debt is reported only as warnings:

python3 scripts/validate_skills.py

Reference files in the repo: STYLE.md (quality baseline), AGENTS.md (contributor guidance), SOUL.md (operating principles).

Caveats

• Authorized testing only — the page repeats this: test only targets you own or have explicit written permission to test.
• Commands assume standard Linux tooling unless a skill says otherwise. Tool availability, scope, network policy, concurrency, credentials, and isolation are all left to the operator.
• Quality is uneven: older skills are still being migrated to the full baseline, which is why style debt surfaces as warnings rather than blocking errors.