CF ssrf
Hunt Server-Side Request Forgery (CWE-918) — any outbound fetch whose URL, scheme, host or port derives from a request. Activate on avatar-by-URL, webhook/callback registration, URL preview/unfurl, render-to-PDF/screenshot/thumbnailer, import-from-URL, feed ingest, OAuth redirect_uri/jwks_uri discovery, "test connection" probes, proxy and media passthrough, and on helpers named download/fetch/get_image/get_base64_image_blob/load_url whose bodies you cannot see. Ships probe.sh, which greps fetch sinks across seven ecosystems, url-parameter contracts, their call sites and allow-list posture in one pass. Keywords SSRF, requests.get, urlopen, httpx, fetch, axios, OkHttp, curl_exec, 169.254.169.254, cloud metadata, gopher://, allow_redirects, DNS rebinding, blind SSRF, URLField.
This skill hunts SSRF vulnerabilities — when an app makes requests using URLs from user input. Promises to find fetch calls across seven ecosystems and check if they're guarded against internal addresses.
Two files and probe.sh script that greps keywords (requests.get, httpx, fetch, curl_exec, etc.). Never ran on models. Code quality scores 65, process scores 28. No critical findings logged, no validation results shown.
Works like grep on source code, not dataflow analysis. Catches obvious cases, misses indirect calls and complex chains. Good for quick scanning, not for thorough audit.
Hunt Server-Side Request Forgery (CWE-918) — any outbound fetch whose URL, scheme, host or port derives from a request.
As a process F 28/100 · No process to follow — weak spots: steps, result and completion, when it triggers
How to improve
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- 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 · 0
✓ No critical or high findings
Files scanned: 2. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-no-whendescription does not say WHEN to use the skill (no "use when")
Process rating: all ten parameters 28/100
- 0Steps. Prose only: no discrete steps
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 3 mutating operations with no state check
- 60Tools and files. Uses tools (web) that frontmatter does not declare
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1334 tokens
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)
- +3No numbered steps or checklist
- +3Output format is not stated: the model decides each time
- +4No input/output examples
- +1No license
- +2Single-language instructions
- +3Description length 784: enough signal without eating the budget
- +4Structure: 6 headings
Quality base 70; lint remarks subtract, signals add up to 100. Result: 65.