BB ag9
Use AG9 to register and verify AI agents with VeryAI Palm-backed human ownership, generate or load portable Ed25519 identities for OpenClaw, Codex, local CLI/MCP, browser, or cloud agents, call AG9 registration/signature verification APIs, and solve reverse-CAPTCHA capability challenges.
As a process B 65/100 · Nearly there — weak spots: result and completion, inputs and preconditions, progress reporting
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".
If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 2
✓ No critical or high findings
Medium and low: 2
-
medium Exfiltration
net-credential-useSKILL.md:288Credential used in a network call (verify the destination is the intended service)curl "https://api.ag9.ai/v1/agent/verify/public-key/$(printf '%s' "$PUBKEY_B64" | jq -sRr @uri)"
-
low Secrets in code
secret-private-keySKILL.md:89Private key material (placeholder value)"privateKeyPem": "-----BEGIN PRIVATE KEY----- …\n",
placeholder
Files scanned: 4. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 5253 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 65/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 50When it triggers. No condition that starts the skill
- 70Execution cost. Instruction body is 5253 tokens
- 100Tools and files. No external tools needed
- 100Steps. 50 steps
- 100Failures and branches. 3 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 10 top-level sections: this looks like several domains in one skill
- low The response is described with custom markup (6 tags): a typed call is more reliable
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Output format is not stated: the model decides each time
- -2localhost URLs: will not work for another user
- +2Single-language instructions
- +3Description length 288: enough signal without eating the budget
- +4Structure: 32 headings
- +3Step-by-step instructions: 50 items
- +4Has examples (17 code blocks)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 73.