Install
$ agentstack add skill-fredoliveira-agents-extract-insights ✓ 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
Extract insights from a note
Purpose
Read one note (transcript, article, clipping, meeting notes — anything) and pull out what's worth keeping, as bullet lists under four headings:
- Actionable lessons — things to do differently, stated as guidance.
- Key insights — the non-obvious ideas worth remembering.
- Todos — concrete next actions that emerge for the reader/user.
- Best practices — repeatable rules or standards worth adopting.
Instructions
- Read the note in full. Take the path the user gave (default to the file
they @-referenced or the active note). Page through long files — don't extract from a partial read.
- Extract. For each of the four headings, write a bullet list. Rules:
- Omit a heading entirely if the note genuinely has nothing for it. Don't pad.
- One idea per bullet, declarative and self-contained (readable without the
source open).
- No fixed count necessary. Capture what's real, skip sponsor reads, filler, and
biography.
- Quote a memorable line only when it earns its place.
- Output — composable, decided by how you were asked:
- Default (no output file mentioned): print the bullet lists directly in
the chat. Nothing is written to disk.
- If the user asks for a markdown file (e.g. "save it", "create a note",
or gives a path): write the lists to a .md file.
- Title / filename:
Insights from .md— keep it
consistent so these notes are easy to spot.
- Location: if the user gave a path, use it. Otherwise default to an
Inbox/ folder at the vault root if one exists (so the user can triage it later), falling back to the source note's own folder.
- Body: link back to the source note in the first line, then the lists.
- Include light frontmatter:
```markdown --- source: "[[]]" created: tags:
- insights
---
# Insights from
Extracted from [[]].
## Actionable lessons
- ...
## Key insights
- ...
``` (Drop empty sections here too.) Then tell the user the path you wrote.
Notes
- Keep it minimal. The model already knows how to extract — this skill just fixes
the four buckets, the bar for inclusion, and the chat-vs-file output choice.
- Adapt to the vault: if the user's notes use different tags or folders, match them.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: fredoliveira
- Source: fredoliveira/agents
- 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.