CAUTION version 1.0.0 · audited 2026-09-09
This is a documentation-only skill: it is a Markdown cheat sheet showing curl commands for the official Trello REST API. It tells the agent to read two secrets from environment variables (TRELLO_API_KEY and TRELLO_TOKEN) and pass them as query parameters to api.trello.com, the legitimate Trello endpoint. Commands cover listing boards/lists/cards and creating, moving, commenting on, and archiving cards — all consistent with the stated purpose. There are no scripts, no downloads, no file writes, no persistence, no obfuscated code, and no third-party destinations. The only real risks are inherent: the Trello token grants full account access, and the write commands (create/move/comment/archive) can change board data if an agent runs them without confirmation.
| Declared purpose | Manage Trello boards, lists, and cards via the Trello REST API. |
| Observed behavior | Provides shell (curl + jq) snippets that call only https://api.trello.com using TRELLO_API_KEY and TRELLO_TOKEN from the environment; read operations (boards, lists, cards) and write operations (create card, move card, comment, archive). No executable files, no network destination other than Trello, no data written to disk. |
| Verdict, rules only | CAUTION · reads credential-like environment variables: TRELLO_API_KEY, TRELLO_TOKEN |
| Verdict, AI | SAFE |
| Final verdict | CAUTION (never better than either pass) |
| capability | vs purpose | detail |
|---|---|---|
| network | needed | All calls go to https://api.trello.com/1/... (SKILL.md lines 28-83). No other hosts are contacted; clawhub.ai and developer.atlassian.com appear only as documentation links. |
| shell | needed | The skill is a set of bash/curl/jq command examples the agent is expected to run (SKILL.md line 17, 24). |
| credentials | needed | Reads TRELLO_API_KEY and TRELLO_TOKEN from environment and sends them as URL query parameters to Trello only. The token grants full Trello account access (noted at line 70). |
| severity | finding |
|---|---|
| info | Secrets passed in URL query string to the official Trello API The API key and token are placed in the URL rather than a header. This is how Trello's API works, and the destination is the legitimate api.trello.com, but URLs with secrets can end up in shell history, process listings, or proxy logs. SKILL.md:28 curl -s "https://api.trello.com/1/members/me/boards?key=$TRELLO_API_KEY&token=$TRELLO_TOKEN" | jq '.[] | {name, id}' |
| low | Credential grants broad account access The skill itself discloses that the supplied token has full access to the user's Trello account. Any agent using this skill can read and modify all boards the token can reach. SKILL.md:70 - The API key and token provide full access to your Trello account - keep them secret! |
| low | Destructive write operations included Documented commands can archive cards, move cards between lists, and post comments. These are within the declared purpose but change user data; an agent acting autonomously could modify boards without confirmation. SKILL.md:63 curl -s -X PUT "https://api.trello.com/1/cards/{cardId}?key=$TRELLO_API_KEY&token=$TRELLO_TOKEN" \ |
0 AI finding(s) were dropped because their file, line or quote did not match the package.
| domain | service rating | where |
|---|---|---|
| api.trello.com | not in the service index | SKILL.md:28 |
| clawhub.ai | not in the service index | skill-card.md:7 |
| developer.atlassian.com | not in the service index | SKILL.md:4 |
| trello.com | not in the service index | SKILL.md:14 |
| Binaries invoked | bash, curl |
| Environment variables read | TRELLO_API_KEY, TRELLO_TOKEN |
| Hard-coded wallets | none |
| Pipes a download to a shell | 0 |
| eval / exec / subprocess | 0 |
| base64 blobs | 0 |
| File writes | 0 |
| Persistence | 0 |
| file | lines | sha256 |
|---|---|---|
SKILL.md | 84 | ede0f3181af05734… |
_meta.json | 6 | fc85bd1c083d1149… |
skill-card.md | 42 | f530015a0ab2fea2… |
For agents
JSON: https://agenteconomy.report/k/trello.audit.json · badge: https://agenteconomy.report/k/trello.audited.svg ·
skill rating: /k/trello · commission an audit of any skill: US$ 29 per version.
The complete published package of this exact version was downloaded from the registry and read statically; nothing was executed. A deterministic pass extracts network destinations, binaries, environment variables, writes, install commands, obfuscation markers, subprocesses and wallets, each with file and line. An AI then reads the whole package with those facts and writes the summary, the capabilities and the findings under a strict schema; every finding it produces must cite a file, a line and the exact text, or it is dropped. The final verdict is the worst of the two passes. The audit does not change the skill's trust tier (policy); the author may respond through the dispute channel and the response is published here. Commissioned by: the Agent Economy Report (free program: the 150 most downloaded skills, September 2026).