More screenshots
Works with the AI Agents you already use
Dependency Bot Config Check
Check Dependabot and Renovate before you trust them: registry tokens in the config, plain-HTTP registries, config the bot rejects, auto-merge workflows that trust github.actor, and automerge with no release-age wait.
$5
Dependency Bot Config Check
Example session with this skill installed
Check this project with dependency-bot-config-check.
- Read your context and instructions
- Compiled the dependency bot config
$ python3 scripts/check_depbot.py examples/risky
BLOCK api-service/renovate.json:13 D01 LITERAL_REGISTRY_CREDENTIAL
"token": "<redacted>"
Why: A registry password, token or key is written in the bot's config (a Dependabot registries entry, a Renovate hostRules entry or header, npmToken, an _authToken line in npmrc, or a password in the user-info part of a registry URL). Everyone who can read the repository, its forks or its history can use it to read, and often publish, your private packages.
Safer: Store it as a Dependabot secret and write ${{secrets.NAME}}, or as a Renovate secret (Mend app settings or self-hosted secrets) and write {{ secrets.NAME }}. Revoke and replace the credential that was committed.
BLOCK api-service/renovate.json:19 D02 PLAIN_HTTP_REGISTRY
BLOCK web-app/.github/dependabot.yml:6 D02 PLAIN_HTTP_REGISTRY
BLOCK web-app/.github/dependabot.yml:7 D01 LITERAL_REGISTRY_CREDENTIAL
BLOCK web-app/.github/dependabot.yml:26 D03 CONFIG_BOT_CANNOT_USE
BLOCK web-app/.github/workflows/dependabot-auto-merge.yml:11 D04 BOT_CHECK_BY_ACTOR
REVIEW api-service/renovate.json:3 D16 UNPINNED_REMOTE_PRESET
REVIEW api-service/renovate.json:4 D11 AUTOMERGE_WITHOUT_RELEASE_AGE
REVIEW api-service/renovate.json:4 D12 AUTOMERGE_ALL_UPDATE_TYPES
REVIEW api-service/renovate.json:5 D18 MERGE_WITHOUT_CHECKS
REVIEW api-service/renovate.json:7 D15 VULNERABILITY_ALERTS_OFF
REVIEW web-app/.github/dependabot.yml:16 D11 AUTOMERGE_WITHOUT_RELEASE_AGE
REVIEW web-app/.github/dependabot.yml:18 D13 MAJOR_UPDATES_IGNORED
REVIEW web-app/.github/dependabot.yml:22 D17 EXTERNAL_CODE_EXECUTION_ALLOWED
REVIEW web-app/.github/dependabot.yml:23 D14 UPDATES_TURNED_OFF
REVIEW web-app/.github/dependabot.yml:27 D19 UPDATE_DIRECTORY_MISSING
REVIEW web-app/.github/workflows/dependabot-auto-merge.yml:19 D12 AUTOMERGE_ALL_UPDATE_TYPES
REVIEW web-app/.github/workflows/dependabot-auto-merge.yml:19 D18 MERGE_WITHOUT_CHECKS
(Each finding also has Why and Safer lines, as in the first one.)
Verdict: BLOCK. Checked 4 files, 3 with dependency bot content.
This is a static check of file text. NOT FLAGGED is not a guarantee.
Connects securely to your tools. The creator never sees your data.
About this skill
An auto-merge workflow on pull_request_target that checks if: github.actor == 'dependabot[bot]' can merge an attacker's pull request, because github.actor is whoever caused the last event, not who opened the pull request. A registry token pasted into dependabot.yml or renovate.json is readable by everyone with the repository. A dependabot.yml with package-ecosystem: "yarn" is rejected, and the updates you expect never arrive. This skill finds those and the other dependency bot mistakes below, and tells you how to fix each one.
Your agent runs a scanner on .github/dependabot.yml, Renovate config files (renovate.json, renovate.jsonc, renovate.json5, .renovaterc files, also under .github/ and .gitlab/), the renovate key of package.json, and the workflows in .github/workflows/ that approve or merge Dependabot or Renovate pull requests (other workflows are skipped). It reads JSON with comments and JSON5 the way Renovate does, treats ${{secrets.NAME}} and {{ secrets.NAME }} as secrets kept outside the file, knows Dependabot's 3-day default cooldown, and checks that each Dependabot directory exists in the repository. It never runs the bots or a workflow and never uses the network. It returns BLOCK, REVIEW, or NOT FLAGGED, with the file, line, reason, and a safer approach. Secrets are never printed.
What it checks
BLOCK
- D01: registry passwords, tokens and keys written in the config (Dependabot registries, Renovate hostRules and auth headers, npmToken, _authToken in npmrc, a password inside a registry URL)
- D02: registries reached over plain HTTP, or Renovate insecureRegistry: true
- D03: config the bot cannot use: invalid Renovate JSON, tab indentation, missing version: 2, updates, package-ecosystem, directory or schedule.interval, wrong ecosystem or interval names, undefined registries or groups, dependabot.yml outside .github/
- D04: workflows on pull_request_target or workflow_run that approve or merge bot pull requests based on github.actor or the event sender
REVIEW
- D11: automerge with no release-age wait (Renovate minimumReleaseAge unset, or Dependabot cooldown 0 days with an auto-merge workflow)
- D12: every update type merged on its own, major versions included
- D13: major updates ignored for every dependency
- D14: updates switched off (open-pull-requests-limit: 0, ignore "*" for everything, Renovate enabled: false)
- D15: Renovate vulnerability alerts turned off
- D16: remote Renovate presets not pinned to a tag
- D17: Dependabot insecure-external-code-execution: allow
- D18: merges that skip checks (gh pr merge without --auto or with --admin, direct merge API calls, Renovate ignoreTests with automerge)
- D19: Dependabot directories that do not exist or hold no manifest for their ecosystem
What's included
- The scanner (Python 3.9+, standard library only)
- A rules guide with the reason, safer approach and official source for each rule
- A risky and a safe example (a web app with Dependabot and an auto-merge workflow, and a service with Renovate), and the real report the scanner produces on the risky one
Limits It reads each file as written. Settings made outside the repository (bot settings in the GitHub or Mend UI, Dependabot secrets, branch protection, GitHub's auto-merge setting) are not seen, and remote presets are named but not fetched, so their content is not checked. Workflows are read only for the bot merge patterns above; general workflow security such as script injection or token permissions is not covered. NOT FLAGGED is not a guarantee.
Static check of config and workflow text. Bot settings made in the GitHub or Mend UI, the content of remote presets, branch protection and general workflow security are not checked. NOT FLAGGED is not a guarantee.
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 5 days ago
Requires
Permissions
Runs one local Python script in the terminal. It reads your project's files, writes nothing, and makes no network calls. The example projects include clearly fake, inert registry tokens (not-a-real-... NOTREAL) for the scanner to find. Nothing in them is run.