voice-learn
Improves a voice profile by learning from manual edits. Use after editing generated text to refine registers and close voice drift over time.
How do I install this agent skill?
npx skills add https://github.com/athola/claude-night-market --skill voice-learnIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The skill is generally safe and functions as intended to learn voice preferences. It carries a low risk of indirect prompt injection as it processes user-provided text without strong isolation, and it manages files within the user's home directory.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
Voice Learning Skill
Learn from user edits to improve the voice profile over time.
When NOT To Use
- Building the first profile (use
scribe:voice-extract) - Reviewing text without changing the profile (use
scribe:voice-review)
Method: Three-Stage Comparison
Every piece flows through three stages:
- Pre-review: Raw generation output (before review agents)
- Post-review: After user accepts/rejects advisory fixes
- Post-edit: User's manually edited final version
The learning agent compares stages 2 and 3 (post-review vs post-edit) to identify patterns in what the user changed. These patterns inform register and rule updates.
Core Rules
- Sharpen, don't add: Modify existing rules to cover new patterns. Rule bloat degrades output.
- Tag specificity: Register-specific patterns go to registers. Universal patterns go to craft rules or agents.
- Flag contradictions: Opposite patterns across pieces require user resolution.
- Evidence threshold: Patterns need 3+ instances (or 1-2 matching existing accumulator entries) before becoming rules.
- Detection surface: Structural changes increase AI detectability. Craft-level changes are neutral. Prefer craft-level updates.
- Rule count check: Suggest consolidation if any section has 8+ rules.
Required Progress Items
voice-learn:snapshots-loaded- All three stages readvoice-learn:diff-analyzed- Changes categorizedvoice-learn:accumulator-checked- Prior patterns reviewedvoice-learn:proposals-generated- Updates proposedvoice-learn:user-approved- Changes accepted by user
Step 1: Load Snapshots
Load: @modules/snapshot-management
PROFILE_DIR="$HOME/.claude/voice-profiles/{name}"
SNAP_DIR="$PROFILE_DIR/learning/snapshots"
# Find the most recent snapshot set
# Format: {piece-name}-{timestamp}-{stage}.md
Read all three stages for the target piece.
Steps 2-4 need Opus or larger: telling a tone adjustment from a precision edit takes the nuance Sonnet flattens. When this session runs Sonnet or Haiku, run them as one dispatch:
Agent(general-purpose):
- model: opus
- prompt: {Steps 2-4 of this skill, the three snapshot paths,
and learning/accumulator.json}
The subagent sees nothing from this conversation and reads the snapshots from disk. It returns the proposals and does not write the profile. Step 5 approval stays in this session. On an Opus or Fable session, run the steps inline.
Step 2: Diff Analysis
Load: @modules/pattern-analysis
Compare post-review vs post-edit. Categorize every change:
| Category | Example |
|---|---|
| Tone adjustment | Softened a claim, added hedge |
| Voice insertion | Added parenthetical, aside, humor |
| Structure change | Broke paragraph, reordered |
| Precision edit | Replaced vague with specific |
| Deletion | Removed fluff or decoration |
| Addition | Added context, example, anchor |
Step 3: Check Accumulator
Read learning/accumulator.json:
{
"patterns": [
{
"id": "pat-001",
"category": "tone_adjustment",
"description": "Softens confident claims about tool capabilities",
"instances": [
{"piece": "blog-post-1", "date": "2026-04-08", "diff": "..."}
],
"target": "register",
"status": "accumulating",
"first_seen": "2026-04-08",
"last_seen": "2026-04-08"
}
],
"staleness_threshold_days": 30
}
Match new changes against existing patterns:
- Semantic similarity (same category + similar description)
- If match found: merge instance, check if threshold reached
- If no match: create new accumulator entry
Step 4: Generate Proposals
For patterns that reach threshold (3+ instances or 1-2 matching prior accumulator entries with 2+ instances):
Apply (strong evidence)
## Proposed Update
**Pattern**: {description}
**Target**: {register file or craft-rules.md}
**Evidence**: {N instances across M pieces}
| Piece | Date | Change Made |
|-------|------|-------------|
| ... | ... | ... |
**Proposed edit**:
- File: {path}
- Section: {section name}
- Current: "{current text or 'new addition'}"
- Proposed: "{new text}"
Hold (insufficient evidence)
Add to accumulator with current instances. Report:
Holding: "{pattern description}" (N instances, need 3+)
Contradictions
If a new pattern contradicts an existing accumulator entry:
Contradiction detected:
- Existing: "{accumulator pattern}"
- New: "{contradicting pattern}"
- Resolution required: user must choose
Step 5: User Approval
Present proposals to user:
Learning found N patterns ready to apply:
[1] {pattern}: {proposed change}
Evidence: {N instances}
[a]pply / [s]kip / [v]iew evidence?
[2] ...
Apply approved changes to the target files.
Staleness
Patterns in the accumulator expire after staleness_threshold_days
(default 30). If a pattern hasn't recurred within that window,
it was likely a one-off preference rather than a voice trait.
On each learning pass, prune stale entries:
# Remove patterns older than threshold with < 3 instances
Snapshot Capture
The learning system captures snapshots automatically when voice-review completes. Snapshot naming:
{piece-filename}-{YYYYMMDD-HHMMSS}-pre-review.md
{piece-filename}-{YYYYMMDD-HHMMSS}-post-review.md
{piece-filename}-{YYYYMMDD-HHMMSS}-post-edit.md
The post-edit snapshot is captured when the user runs
/voice-learn after finishing their manual edits.
Exit Criteria
- Snapshots loaded and compared
- Changes categorized
- Accumulator checked and updated
- Proposals generated for threshold patterns
- User approved/rejected proposals
- Approved changes applied to profile files
- Stale accumulator entries pruned
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/athola/claude-night-market/voice-learn">View voice-learn on skillZs</a>