Install
$ agentstack add skill-nearform-unwind-uw-verify ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo 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.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
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 →About
Verifying Layer Documentation
Purpose: Detect gaps between documentation and source code. Output is a work list for the next skill to fix.
This is now a deterministic check. Completeness is verified by set arithmetic (manifest − docs), not a subjective LLM comparison. The scanner already knows the ground-truth set of items per layer; this skill diffs that against what the docs actually contain (matched by anchor id, then by name).
Output:
docs/unwind/.cache/coverage/{layer}.json- per-layer coverage reportdocs/unwind/layers/{layer}/gaps.md- missing items work list (only when gaps exist)
Process
Step 1: Run the Coverage Verifier
# Locate the installed Unwind plugin, then load the core helper.
# $0/BASH_SOURCE are unreliable under `bash -c`, so glob the install cache.
UNWIND_PLUGIN_ROOT="${CLAUDE_PLUGIN_ROOT:-${UNWIND_PLUGIN_ROOT:-}}"
[ -f "$UNWIND_PLUGIN_ROOT/skills/scripts/_resolve-plugin-root.sh" ] || \
UNWIND_PLUGIN_ROOT="$(ls -dt "$HOME"/.claude/plugins/cache/*/unwind/*/ 2>/dev/null | head -1)"
source "${UNWIND_PLUGIN_ROOT%/}/skills/scripts/_resolve-plugin-root.sh"
ensure_unwind_core || echo "core unavailable — using legacy LLM gap detection"
node "$UNWIND_PLUGIN_ROOT/skills/scripts/verify-coverage.mjs" "$(pwd)"
The script:
- Loads
docs/unwind/.cache/scan-manifest.json(the ground-truth candidate set). - Reads every section
.mdunder eachdocs/unwind/layers/{layer}/folder and
extracts documented item ids (from `` anchors, then fuzzy by name).
- Diffs them: missing = in source, not in docs; extra = documented but
not in scan (flagged for review — may be dynamic or stale).
- Writes
coverage/{layer}.jsonand agaps.mdwork list for any layer with
missing items.
Step 2: Summarize
Report the per-layer coverage table the script prints (covered/total = pct%), and point the user/orchestrator at the generated gaps.md files. Then hand off to uw-complete to fill them.
Output Format (generated gaps.md)
# {layer} Documentation Gaps
> Generated by verify-coverage. Coverage: 36/37 (97.3%).
## Missing Items
### users
- **Type:** table
- **Location:** src/db/schema.ts:12-34
- **Id:** `table:src/db/schema.ts:users`
uw-complete reads this verbatim, documents each item under the correct section using the anchor-id heading format, and deletes gaps.md.
Legacy Fallback (no @unwind/core)
When the deterministic core is unavailable (no Node/pnpm, or build failed), fall back to the subjective comparison: for each layer, read the docs and the source entry points from architecture.md, build the two item lists by hand, and write the gaps the LLM finds to docs/unwind/layers/{layer}/gaps.md in the same format. This is slower and not reproducible — it exists only so the plugin still functions without the deterministic engine.
Next Step — continue or pause?
The branch depends on the coverage table just printed:
- Any layer Pipeline: scan → analyze → verify ✓** → complete → plan → dashboard.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: nearform
- Source: nearform/unwind
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.