cti-setup

In-chat configuration of API keys for the threat-intel integrations bundled with this pack. Use this when the user is in Claude Code and doesn't want to drop to a shell to run ./scripts/setup.sh.

When to invoke

• User asks "how do I set up keys", "configure my API keys", "add a VirusTotal key", etc.
• User runs /cti-setup
• A lookup-* skill failed because a key is missing and you want to offer to add it
• After install via /plugin marketplace add or npx (no shell setup ran)

What you do

1. Check current state. Read .claude/settings.local.json. If it's missing or has no env block, the user has zero keys configured. If it has some, list which are present and which are missing.
2. Tell the user the menu. Present the services in a table with: name, env variable, free-tier limit, signup URL. Make clear all are optional and that the pack degrades gracefully.
3. Ask which to configure. Let the user provide one, several, or all. Don't force them through every prompt.
4. Receive the keys. When the user shares a key, treat it as sensitive — do not echo it back in plain text in your response (refer to it as your VirusTotal key or the masked tail …<last 4 chars>).
5. Write the merged file. Use the non-destructive merge below — preserve every other field in settings.local.json.
6. Offer to verify. Ask if they want you to dry-run each configured key against its CLI to confirm it's wired up.
7. Tell them what's next. "Try /ip-investigation 8.8.8.8" or similar concrete next command.

The services

If the user has a Liberty91 account, configure LIBERTY91_API_KEY first — it is the pack's first-party source and /lookup-liberty91 runs before third-party lookups, so it saves other services' quota. Keys are l91_live_ (production) or l91_test_ (development); an empty scope list on the key grants all read scopes, which is the right default for enrichment. Set LIBERTY91_API_URL only to point at a non-production host.

A starter set of VirusTotal + OTX + URLScan + AbuseIPDB covers most IP/domain/URL/hash investigations. Shodan and GreyNoise add value for IP-focused work. Censys is optional (very tight rate limit). MISP requires both a base URL and an auth key — point it at your org's instance. OpenCTI likewise takes a base URL plus an API token and powers /lookup-opencti (two-way: query your knowledge base + push vetted intel back). Ransomware.live powers the lookup-ransomwarelive and ransomware-ecosystem skills (victim/group tracking). ReversingLabs is a licensed product — only configure if your organisation has a Spectra Analyze (A1000) account. CrowdStrike Falcon Intelligence is a licensed subscription — it powers /lookup-crowdstrike for IOC reputation AND threat-actor / TTP / report intelligence; configure if your org has a Falcon Intelligence licence with Intel API scopes. Microsoft Sentinel takes four values and powers /lookup-sentinel (hunt your own workspace: IOC exposure sweeps + ATT&CK TTP hunts, read-only). All four come from the Azure portal: create an Entra ID app registration (→ tenant id + client id), add a client secret (shown once), grant the app Log Analytics Reader on the Sentinel workspace, and copy the Workspace ID from the workspace Overview blade — the step-by-step is in tools/integrations/sentinel.md. No extra licence is needed beyond the workspace itself.

How to write the file

The file is .claude/settings.local.json. It is gitignored. Do not overwrite it — read, merge the env block, write back. Use the bundled setup script which handles this safely:

./scripts/setup.sh --non-interactive \
  --liberty91=USER_PROVIDED_KEY \
  --virustotal=USER_PROVIDED_KEY \
  --shodan=USER_PROVIDED_KEY \
  --misp-url=https://misp.example.org \
  --misp=USER_PROVIDED_KEY \
  --opencti-url=https://opencti.example.org \
  --opencti=USER_PROVIDED_TOKEN \
  --ransomwarelive=USER_PROVIDED_KEY \
  --reversinglabs-user=USER_PROVIDED_USERNAME \
  --reversinglabs-password=USER_PROVIDED_PASSWORD \
  --reversinglabs-host=https://a1000.reversinglabs.com

(Pass only the flags for keys the user actually shared. Available flags: --liberty91, --liberty91-url, --virustotal, --urlscan, --shodan, --abuseipdb, --greynoise, --otx, --censys, --misp-url, --misp, --opencti-url, --opencti, --ransomwarelive, --reversinglabs-user, --reversinglabs-password, --reversinglabs-host, --crowdstrike-client-id, --crowdstrike-client-secret, --crowdstrike-base-url, --sentinel-tenant-id, --sentinel-client-id, --sentinel-client-secret, --sentinel-workspace-id.)

If scripts/setup.sh is not present (e.g. plugin-only install), do the merge yourself with this Node one-liner. Replace KEY=VAL pairs with the user's input:

node -e "
  const fs = require('fs');
  const path = '.claude/settings.local.json';
  let cur = {};
  try { cur = JSON.parse(fs.readFileSync(path, 'utf8')); } catch(e) {}
  cur.env = cur.env || {};
  Object.assign(cur.env, {
    VIRUSTOTAL_API_KEY: 'USER_PROVIDED_KEY',
    SHODAN_API_KEY: 'USER_PROVIDED_KEY',
  });
  fs.mkdirSync('.claude', { recursive: true });
  fs.writeFileSync(path, JSON.stringify(cur, null, 2) + '\n');
"

After writing, confirm to the user:

• which keys were added (by service name, not the key value)
• which keys are still unconfigured
• that the file is gitignored

Verifying keys

If the user wants to verify, run:

./scripts/setup.sh --verify

This dry-runs each lookup CLI and reports OK/fail per service without making a real API call. If setup.sh is unavailable, dry-run each CLI individually:

node tools/clis/virustotal.js ip 8.8.8.8 --dry-run

Exit code 0 = key present and CLI invocation OK. Exit code 2 = missing key.

Removing or rotating a key

To remove a key, edit .claude/settings.local.json and delete the entry from the env block (or set its value to ""). To rotate, just re-run setup with the new value — the merge overwrites that key only.

Security notes

• Never commit .claude/settings.local.json (already gitignored at repo root).
• Never echo the full key value back to the user in chat — they shared it once; mask thereafter.
• Don't write keys to logs, screenshots, or other artefacts.
• If the user accidentally pastes a key into a public channel, advise them to rotate it on the provider's site and re-run /cti-setup with the new value.

What this does NOT do

• Does not call the threat-intel APIs (only --dry-run invocations of the local CLIs).
• Does not download MITRE ATT&CK data — for that, run ./scripts/download-mitre.sh (the /mitre-attack skill self-heals on first use).
• Does not configure permissions, hooks, or other Claude Code settings — only the env block of settings.local.json.