AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Web Evidence

skill-graysurf-agent-kit-web-evidence · by graysurf

Capture redacted static HTTP evidence bundles through the nils-cli web-evidence command.

No reviews yet
0 installs
26 views
0.0% view→install

Install

$ agentstack add skill-graysurf-agent-kit-web-evidence

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-graysurf-agent-kit-web-evidence)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
Archived

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Web Evidence? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Web Evidence

Use this skill when an agent workflow needs deterministic, redacted evidence for one HTTP or HTTPS URL.

Contract

Prereqs:

  • web-evidence available on PATH from nils-cli 0.8.4 or newer.
  • Static HTTP/HTTPS evidence only. This command does not drive a browser, execute JavaScript, reuse cookies, or use authenticated sessions.
  • The caller chooses an explicit output directory, preferably under a project-scoped agent-out run directory.

Inputs:

  • capture: required URL and --out DIR; optional --format text|json, --label LABEL, --method get|head,

--timeout-seconds N, --max-body-bytes N, and --body-preview-bytes N.

  • completion: required shell name bash|zsh.

Outputs:

  • web-evidence capture: writes a redacted artifact bundle under --out DIR.
  • summary.json: versioned summary with request metadata, status classification, artifact paths, redaction counts, and error metadata.
  • headers.redacted.json: request and response header names with sensitive values redacted.
  • body-preview.redacted.txt: truncated, redacted text body preview when the response body is text.
  • JSON stdout with --format json uses schema_version value cli.web-evidence.capture.v1.

Exit codes:

  • 0: evidence captured and response classified as successful.
  • 1: runtime, network, or HTTP status failure; when possible, the command still writes the redacted bundle.
  • 64: usage or configuration error.

Failure modes:

  • web-evidence is unavailable on PATH.
  • URL is unsupported, malformed, or not HTTP/HTTPS.
  • Output directory cannot be created or safely written.
  • Network request fails, times out, or returns a failure HTTP status.
  • Caller needs browser state, JavaScript execution, screenshots, console logs,

cookies, or auth headers; use Browser/Chrome/Playwright tooling instead.

Setup

Released PATH boundary:

web-evidence --help
web-evidence capture --help

Use the PATH command after installing nils-cli 0.8.4 or newer with nils-web-evidence on PATH.

Local checkout fallback boundary:

cargo run --locked --manifest-path "$HOME/Project/sympoies/nils-cli/Cargo.toml" \
  -p nils-web-evidence --bin web-evidence -- --help

Run the Cargo form from the workflow's target directory only when PATH is absent or reports an older nils-cli, and only from a validated local nils-cli checkout. Keep the same web-evidence subcommands and flags in both modes.

Commands (only entrypoints)

Released PATH command:

web-evidence capture  --out  [--format text|json] [--label ] [--method get|head]
web-evidence completion 

Local checkout fallback command:

cargo run --locked --manifest-path /path/to/nils-cli/Cargo.toml \
  -p nils-web-evidence --bin web-evidence --  ...

Do not reimplement HTTP fetching, redaction, schema generation, or artifact naming in skill-local scripts.

Workflow

  1. Create or choose a run directory for the current task, usually with agent-out project --topic --mkdir.
  2. Capture static evidence:

web-evidence capture --out /web-evidence --label --format json

  1. Read summary.json or JSON stdout for ok, status_class, status_code, artifacts, and error.
  2. Attach or cite only redacted artifacts from the bundle; do not preserve raw cookies, auth headers, URL secrets, or unredacted network logs.
  3. If browser behavior is required, report that this static evidence command is insufficient and use a browser tool instead.

References

  • docs/runbooks/skills/TOOLING_INDEX_V2.md

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.