Skill Observer
Review pending skill suggestions collected by the background observer. Use when the user wants to see, approve, reject, or apply pattern suggestions that the observer flagged during past sessions.
Install
$ agentstack add skill-msdakot-ai-foundary-skill-observer ✓ 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.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-msdakot-ai-foundary-skill-observer) 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
Skill Observer — Suggestion Review
The background observer watches conversation turns and flags repeated correction patterns. This skill lets the user review and act on those findings.
Step 1 — Read pending suggestions
cat ~/.claude/observer/pending-suggestions.jsonl 2>/dev/null || echo "No pending suggestions."If empty, tell the user: "No suggestions pending. The observer will flag patterns as they accumulate across sessions."
Step 2 — Present each finding clearly
For each entry in the file, show the user:
┌─ Suggestion ─────────────────────────────────────────────────────┐
│ Type: │
│ Theme: │
│ Confidence: │
│ Evidence: │
│ │
│ Proposed: │
└──────────────────────────────────────────────────────────────────┘Ask: "Apply this suggestion? [y]es / [n]o / [e]dit"
Step 3 — Act on user response
y — approve:
claude-md: append the suggestion text as a bullet to~/.claude/CLAUDE.md. Create the file if it doesn't exist.new-skill: create a new SKILL.md at~/.claude/skills//SKILL.mdwith the suggestion as the body. Tell the user to register it in their settings if needed.patch-skill: show the user which skill file to edit, display the suggested change, and apply it if they confirm.
n — reject:
- Mark dismissed. Do not apply.
e — edit:
- Show the suggestion text in full, let the user dictate changes, then apply the edited version.
Step 4 — Clear applied/rejected suggestions
After processing all suggestions, remove handled entries from the pending file:
# Clear the file after all suggestions are resolved
> ~/.claude/observer/pending-suggestions.jsonlTell the user how many were applied vs dismissed.
Notes
- Low-confidence suggestions: flag these explicitly and recommend the user scrutinize before approving.
- If a
patch-skillsuggestion names a skill that doesn't exist, treat it asnew-skillinstead. - Never auto-apply without explicit user approval.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: msdakot
- Source: msdakot/ai-foundary
- 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.