- Home
- Skills
- Agents & Orchestration
- Takeover Bleeding Check
Works with the AI Agents you already use
Takeover Bleeding Check
Just took over a project? Find what's leaking today: exposed keys, forgeable logins, open routes, leftover dev tools, missing backups.
$5
Takeover Bleeding Check
Example session with this skill installed
I just took over a courier-booking back office from a vendor who left. Before I touch anything, run /bleeding-check.
- Read your context and instructions
- Compiled the takeover bleeding check
(Example output on a fictional project. Full sample in examples/bleeding-check-example.md)
First-hour list
| # | Problem | Evidence | Action |
|---|---|---|---|
| 1 | Sessions can be forged and never expire: cookie is base64(SECRET:timestamp), so one leaked cookie reveals the secret | src/lib/auth.ts:4,30-38 | Server-side random sessions with expiry; logout must revoke |
| 2 | Public endpoint returns customer name, phone, address; ids are date + 4 random chars, enumerable | api/public/booking/[id]/route.ts:7-12 | Return QR + status only; unguessable tokens; rate limit |
| 3 | Dev proxy signing real carrier requests runs on production port 3001; its token is printed in the README | proxy/proxy.js:10, README.md:117 | See human check 1 |
| 4 | Admin over plain HTTP; basic-auth layer removed | README.md:78 | HTTPS; IP allowlist meanwhile |
| 5 | No database backup anywhere | deploy.sh, crontab | Daily off-server backup, then restore once |
Human checks
- Is port 3001 open in the cloud firewall? If yes, stop the proxy today.
- Whose card pays for the LLM API account?
- Is 203.0.113.10 a static IP the client owns? The carrier allowlist is bound to it.
Keys & secrets (excerpt)
| Item | Where | Unlocks | Call |
|---|---|---|---|
| Carrier common key | .env; was hard-coded, still in git history (commit 4b8b904) | Real shipments | Keep (carrier-issued); keep old repo copies private |
| Session secret | .env | See #1 | Rotate after fixing #1 |
| Proxy token | proxy.js + README | See #3 | Stop |
Connects securely to your tools. The creator never sees your data.
About this skill
You just inherited a codebase. Before you spend a week understanding it, find out what is bleeding right now: the things that cost money, leak customer data, or lock you out while you're still reading.
One sitting, one report, a ranked first-hour list.
What it checks
- Keys & secrets: tracked env files, secrets in git history, hard-coded fallbacks (
process.env.X || 'literal'), keys pasted into READMEs and curl examples, non-public keys shipped to the browser. - Open doors: how login actually works (forgeable tokens, no expiry, logout that doesn't revoke), every route that skips auth and what it returns, guessable ids that expose personal data, handlers protected only by middleware.
- Leftovers on production: debug proxies, signing helpers, test endpoints and admin consoles that the deploy script still starts.
- Third-party services: who owns each account, who pays, what's unused but still billing.
- Servers & data: account and domain ownership, IP-bound allowlists, and the classic: no backup, never restored.
Every finding cites a file, line or commit and gets exactly one call: keep / rotate / transfer / stop. Anything the repo can't tell you (firewall rules, who pays the bill) becomes a concrete human-check question.
Field-tested on a production back office, where it found a forgeable session, a public endpoint leaking customer phone numbers and addresses, and a production signing proxy left running. All three had survived a previous cleanup.
Safe by design
- Read-only, no network calls.
- Secrets reported by location only, never printed, not even masked.
- Writes one report file, after asking.
Need the full takeover (scope, rescue-or-rebuild verdict, tests, dependency rescan, first-month monitoring)? See Rescue Audit by the same author.
Works with any coding agent that can read files and run shell commands: Claude Code, Codex CLI, Gemini CLI, Cursor and others.
Can't see cloud consoles, firewalls or billing; those become human-check questions. Doesn't cover requirements, tests or dependency upgrades (see Rescue Audit).
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 8 days ago
Permissions
Reads the repository and git history and runs read-only shell commands (git log, grep). Writes one report file, bleeding-check.md, after asking. No network calls.