anthropics/defending-code-reference-harness

quickstart

- The front door for this repo.

Quelltext ansehen

Verwendung

  1. Installationsbefehl kopieren und das Quellpaket vor der Installation in der eigenen Agentenumgebung prüfen.
  2. Benötigte Werkzeuge, Zugangsdaten und Hosts im Original prüfen.
  3. Eingabe und gewünschtes Ergebnis beschreiben und die Ausgabe mit den Quellen abgleichen.
npx skills add https://github.com/anthropics/defending-code-reference-harness --skill quickstart

Vor dem Start

  • Die Installation umfasst weder Modellabo noch API-Kontingent oder Dateizugriffsrechte.
  • Sterne und Installationen zeigen Aktivität, keine unabhängige Qualitäts- oder Sicherheitsprüfung.
Originales Skill-Dokument

Aus dem Quell-Repository gerendert; Überschriften, Beispiele, Code, Tabellen, Links und Bilder bleiben erhalten.

/quickstart

Two modes, picked by whether $ARGUMENTS is empty.

  • Empty → Intro mode. Short orientation, then offer the guided first run.
  • Non-empty → Help mode. Treat $ARGUMENTS as the operator's question.

Intro mode

Keep it short and a little warm; this is the first thing a new operator sees.

Say roughly:

Welcome! This repo takes you from finding your first vulnerability to patching at scale, using a set of Claude Code skills and an autonomous pipeline. Two ways in: interactive skills (no setup, safe, start here) and the autonomous pipeline (Docker, scales to hundreds of parallel agents). The ramp-up: | Day 1 | Threat-model + first static scan + triage | | Day 2 | Run the reference pipeline (C/C++) | | Day 3-4 | Customize it for your stack | | Week 2 | Autonomous scanning, triage, and patching | Day-1 goal: threat-model, scan, and triage the bundled canary target. Most teams get there before lunch.

Remind them to export CLAUDE_CODE_SUBAGENT_MODEL=<model-id> so subagents use the same model as the session.

Then AskUserQuestion with three options:

  1. Walk me through Day 1 on the canary (~10 min) → run "Guided first

run" below.

  1. I have a question → ask what it is, then switch to Help mode.
  2. I'll read the README → point at README.md Step 1 and stop.

Guided first run

Runs the three Step-1 skills on targets/canary, pausing after each to show what landed on disk. These only read/write files in the repo; no sandbox needed.

  1. /threat-model bootstrap targets/canary via Task. When done, open

THREAT_MODEL.md, show the focus areas, explain in 2-3 sentences how this steers the scan.

  1. /vuln-scan targets/canary via Task. When done, open

targets/canary/VULN-FINDINGS.md, summarize the count and top 2-3 findings, point at VULN-FINDINGS.json.

  1. /triage targets/canary/VULN-FINDINGS.json via Task. When done, open

TRIAGE.md, explain what changed vs. raw findings (verified, deduped, re-ranked).

Pause for the operator between each (AskUserQuestion); don't barrel through. Close with a one-line recap of the three artifacts on disk, then point at README Step 2 for the execution-verified pipeline. Never run `vuln-pipeline` or anything that executes target code here; that's Step 2 and needs Docker + a sandbox.


Help mode

Answer the operator's question using this repo as ground truth: README, docs/*.md, harness/*.py, dnr_harness/*.py, targets/*/config.yaml, .claude/skills/*. Don't answer from general knowledge when the repo has a specific answer.

Routing map

If the question is about…Read firstThen offer
running the pipelinedocs/pipeline.md, README Step 2the recon / run command
too many findings, triagedocs/triage.md/triage <path>
porting, Java/Go/Rust/etc.docs/customizing.md, README Step 3/customize
safety, sandbox, Dockerdocs/security.mdcite; no action
rate limits, 429, token budgetdocs/pipeline.md: Rate limits, docs/troubleshooting.md#rate-limitscite the numbers
duplicates, dedupdocs/troubleshooting.md#duplicate-findingsknown_bugs: hint
CLI flags, "what does --X do"harness/cli.py (grep the argparse)exact flag + example
which model, subagent pinningdocs/troubleshooting.md: Subagentsthe export line
best practices, promptingdocs/best-practices.md, docs/prompting.mdcite the principle
threat model, attack surface, scopedocs/threat-model.md/threat-model bootstrap <target-dir>
scan, audit, find vulns.claude/skills/vuln-scan/SKILL.md/vuln-scan <target-dir>
"how do I start"README Step 1offer Guided first run
patching, fix, diff, re-attackdocs/patching.md, README Step 4/patch <input>
threat hunting, incident response, logsdocs/detection-response.md/dnr-hunt or /dnr-respond
autonomous D&R, dnrcanarydocs/detection-response.md, targets/dnrcanary/README.mdthe dnr-pipeline run command
binary, embedded, other domainsdocs/other-use-cases.mdcite section
anything elseREADME Table of contentsbest-match doc

Answer format

  1. Direct answer in 2-5 sentences.
  2. > source: the file(s) and section you used.
  3. Next action: one copy-pasteable command or skill invocation, if one

applies. If none does, say so.

  1. If the question is ambiguous, ask one clarifying question; don't guess.

Constraints

  • Never fabricate CLI flags or file paths. If unsure, Grep for it in

harness/cli.py or the target configs and quote what you find.

  • If the repo doesn't answer the question, say so plainly and suggest the

operator open a GitHub issue on this repo.

  • Keep the Q&A dry and cited. Save the warmth for Intro mode.

Verwandte Aufgaben

aus demselben Repository

Weitere Skills

Alle Skills
anthropics
Offiziell

vuln-scan

Static source-code vulnerability scan. Reads a target directory (and THREATMODEL.md if present), spawns parallel review subagents per focus area, and writes VULN-FINDINGS.json + .md for /triage to consume. Read-only — no building, running, or network. For execution-verified crashes, use vuln-pipeline instead. Use when asked to "scan for vulns", "review this code for security issues", "find bugs in ", or as the step between /threat-model and /triage.

Installationen / 1h
2
GitHub Stars
7712
Aktualisiert
6. Aug.
anthropics
Offiziell

threat-model

- Build a threat model for a target codebase. Three modes: "interview" walks an application owner through the four-question framework and produces a threat model from their answers; "bootstrap" derives a threat model from the code plus past vulnerabilities (CVEs, git history, pentest reports) when no owner is available; "bootstrap-then-interview" chains the two when both owner and codebase are present. All write THREATMODEL.md in a shared schema. Use when asked to "threat model", "build a threat

Installationen / 1h
1
GitHub Stars
7546
Aktualisiert
6. Aug.
anthropics
Offiziell

dnr-hunt

Proactive threat hunt over web/application logs — no alert in hand. Profiles the corpus, runs a hypothesis-driven hunt loop with a mandatory written ledger, confirms suspects in source, detonates a local PoC, and writes INCIDENTS.json + INCIDENTREPORT.md. Use when asked to "hunt the logs", "find the campaign", "look for signs of compromise", or "run dnr-hunt". The no-alert entry to the detection & response track; /dnr-respond is the lead-in-hand entry.

Installationen / 1h
2
GitHub Stars
7502
Aktualisiert
6. Aug.
anthropics
Offiziell

patch

Generate candidate fixes for verified security findings. Consumes TRIAGE.json (preferred), VULN-FINDINGS.json, INCIDENTS.json, or a vuln-pipeline results directory. Pipeline input is delegated to the execution-verified vuln-pipeline patch ladder; static-analysis input gets a per-finding patch subagent + independent reviewer and is written as inert diffs for human review. Writes PATCHES/bugNN/{patch.diff,patchresult.json}, PATCHES.md, and PATCHES.json. Use when asked to "fix the findings", "patch

Installationen / 1h
2
GitHub Stars
7502
Aktualisiert
6. Aug.