quickstart

Original：🇺🇸 English

Translated

The front door for this repo. With no argument: a 30-second intro, then an offer to walk you through your first run on the canary target. With a question: answers it from this repo's own docs and source, cites where it looked, and hands you the next command. Use for "how do I…", "why does…", "where is…", "can this…", or just "/quickstart" to get oriented.

10installs

Sourceanthropics/defending-code-reference-harness

Added on2026-07-07

NPX Install

npx skill4agent add anthropics/defending-code-reference-harness quickstart

SKILL.md Content

View Translation Comparison →

/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:

Walk me through Day 1 on the canary (~10 min) → run "Guided first run" below.
I have a question → ask what it is, then switch to Help mode.
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.

```
/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.
```
/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
```
.
```
/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

,

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 first	Then offer
running the pipeline	`docs/pipeline.md` , README Step 2	the `recon` / `run` command
too many findings, triage	`docs/triage.md`	`/triage <path>`
porting, Java/Go/Rust/etc.	`docs/customizing.md` , README Step 3	`/customize`
safety, sandbox, Docker	`docs/security.md`	cite; no action
rate limits, 429, token budget	`docs/pipeline.md` : Rate limits, `docs/troubleshooting.md#rate-limits`	cite the numbers
duplicates, dedup	`docs/troubleshooting.md#duplicate-findings`	`known_bugs:` hint
CLI flags, "what does --X do"	`harness/cli.py` (grep the argparse)	exact flag + example
which model, subagent pinning	`docs/troubleshooting.md` : Subagents	the `export` line
best practices, prompting	`docs/best-practices.md` , `docs/prompting.md`	cite the principle
"how do I start"	README Step 1	offer Guided first run
patching, fix, diff, re-attack	`docs/patching.md` , README Step 4	`/patch <input>`
binary, pentest, other domains	`docs/other-use-cases.md`	cite section
anything else	README Table of contents	best-match doc

Answer format

Direct answer in 2-5 sentences.
```
> source:
```
the file(s) and section you used.
Next action: one copy-pasteable command or skill invocation, if one applies. If none does, say so.
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.