ReweaveSAFE
Claude Code plugin that generates individualized knowledge systems from conversation. You describe how you think and work, have a conversation and get a complete second brain as markdown files you own.
Overview
Claude Code plugin that generates individualized knowledge systems from conversation. You describe how you think and work, have a conversation and get a complete second brain as markdown files you own.
9eccf964e924OBSERVED · 2026-10-07What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
---
name: reweave
description: Update old notes with new connections. The backward pass that /reflect doesn't do. Revisit existing notes that predate newer related content, add connections, sharpen claims, consider splits. Triggers on "/reweave", "/reweave [note]", "update old notes", "backward connections", "revisit notes".
user-invocable: true
allowed-tools: Read, Write, Edit, Grep, Glob, Bash, mcp__qmd__search, mcp__qmd__vector_search, mcp__qmd__deep_search, mcp__qmd__status
context: fork
---
## Runtime Configuration (Step 0 — before any processing)
Read these files to configure domain-specific behavior:
1. **`ops/derivation-manifest.md`** — vocabulary mapping, platform hints
- Use `vocabulary.notes` for the notes folder name
- Use `vocabulary.note` / `vocabulary.note_plural` for note type references
- Use `vocabulary.reweave` for the process verb in output
- Use `vocabulary.topic_map` / `vocabulary.topic_map_plural` for MOC references
- Use `vocabulary.cmd_verify` for the next-phase suggestion
2. **`ops/config.yaml`** — processing depth, pipeline chaining
- `processing.depth`: deep | standard | quick
- `processing.chaining`: manual | suggested | automatic
- `processing.reweave.scope`: related | broad | full
If these files don't exist, use universal defaults.
**Processing depth adaptation:**
| Depth | Reweave Behavior |
|-------|-----------------|
| deep | Full reconsideration. Search extensively for newer related {vocabulary.note_plural}. Consider splits, rewrites, challenges. Evaluate claim sharpening. Multiple search passes. |
| standard | Balanced review. Search semantic neighbors and same-{vocabulary.topic_map} {vocabulary.note_plural}. Add connections, sharpen if needed. |
| quick | Minimal backward pass. Add obvious connections only. No rewrites or splits. |
**Reweave scope:**
| Scope | Behavior |
|-------|----------|
| related | Search {vocabulary.note_plural} directly related to the target (same {vocabulary.topic_map}, semantic neighbors) |
| broad | Search across all {vocabulary.topic_map_plural} and semantic space for potential connections |
| full | Complete review including potential splits, rewrites, and claim challenges |
## EXECUTE NOW
**Target: $ARGUMENTS**
Parse immediately:
- If target contains `[[note name]]` or note name: reweave that specific {vocabulary.note}
- If target contains `--handoff`: output RALPH HANDOFF block at end
- If target is empty: find {vocabulary.note_plural} that most need reweaving (oldest, sparsest, most outdated)
- If target is "recent" or "--since Nd": reweave {vocabulary.note_plural} not touched in N days
- If target is "sparse": find {vocabulary.note_plural} with fewest connections
**Execute these steps:**
1. **Read the target {vocabulary.note} fully** — understand its current claim, connections, and age
2. **Ask the reweave question:** "If I wrote this {vocabulary.note} today, with everything I now know, what would be different?"
3. **If a task file exists** (pipeline execution): read it to see what /reflect discovered. The Reflect section shows which connections were just added and which {vocabulary.topic_map_plural} were updated — this is your starting context for the backward pass.
4. **Search for newer related {vocabulary.note_plural}** — use dual discovery (semantic search + {vocabulary.topic_map} browsing) to find {vocabulary.note_plural} created AFTER the target that should connect
5. **Evaluate what needs changing:**
- Add connections to newer {vocabulary.note_plural} that did not exist when this was written
- Sharpen the claim if understanding has evolved
- Consider splitting if the {vocabulary.note} now covers what should be separate ideas
- Challenge the claim if new evidence contradicts it
- Rewrite prose if understanding is deeper now
6. **Make the changes** — edit the {vocabulary.note} with new connections (inline links with context), improved prose, sharper claim if needed
7. **Update {vocabulary.topic_map_plural}** — if the {vocabulary.note}'s topic membership changed, update relevant {vocabulary.topic_map_plural}
8. **If task file exists:** update the {vocabulary.reweave} section
9. **Report** — structured summary of what changed and why
10. If `--handoff` in target: output RALPH HANDOFF block
**START NOW.** Reference below explains methodology — use to guide, not as output.
---
# Reweave
Revisit old {vocabulary.note_plural} with everything you know today. {vocabulary.note_plural} are living documents — they grow, get rewritten, split apart, sharpen their claims. This is the backward pass that keeps the network alive.
## Philosophy
**{vocabulary.note_plural} are living documents, not finished artifacts.**
A {vocabulary.note} written last month was written with last month's understanding. Since then:
- New {vocabulary.note_plural} exist that relate to it
- Understanding of the topic deepened
- The claim might need sharpening or challenging
- What was one idea might now be three
- Connections that were not obvious then are obvious now
Reweaving is not just "add backward links." It is completely reconsidering the {vocabulary.note} based on current knowledge. Ask: **"If I wrote this {vocabulary.note} today, what would be different?"**
> "The {vocabulary.note} you wrote yesterday is a hypothesis. Today's knowledge is the test."
## What Reweaving Can Do
| Action | When to Do It |
|--------|---------------|
| **Add connections** | Newer {vocabulary.note_plural} exist that should link here |
| **Rewrite content** | Understanding evolved, prose should reflect it |
| **Sharpen the claim** | Title is too vague to be useful |
| **Split the {vocabulary.note}** | Multiple claims bundled together |
| **Challenge the claim** | New evidence contradicts the original |
| **Improve the description** | Better framing emerged |
| **Update examples** | Better illustrations exist now |
Reweaving is NOT just Phase 4 of /reflect applied backward. It is a full reconsideration.
## ITrust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
9eccf964e924full audit observations/trust-audit/skill/agenticnotetaking__reweave.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | 9eccf964e924 | SAFE | B | 89 | first audit |
Questions
What does the Reweave skill do?
Claude Code plugin that generates individualized knowledge systems from conversation. You describe how you think and work, have a conversation and get a complete second brain as markdown files you own.
Is Reweave safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Reweave access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (9eccf964e924), read on 2026-10-07. The repository is watched, and a new audit runs when it changes — this is the first audit.