From design-system-ops
Persists and recalls findings from design system skill runs across sessions for comparison and correlation by other skills.
npx claudepluginhub murphytrueman/design-system-opsThis skill uses the workspace's default tool permissions.
A skill for persisting findings across skill runs so that future skills can compare, correlate, and build on what was previously discovered.
Logs cross-project outcomes and recalls lessons to inform new sessions, avoiding past mistakes. Analyzes skill executions for better routing. Use /memento modes: log, global recall, health, route.
Searches, recovers, and analyzes AI session histories across Claude Code, AI Studio, and Gemini CLI. Useful for recovering lost files after compaction, searching conversations, detecting corrections, and self-improvement workflows.
Manages cross-session handoff, durable learning, and memory persistence by storing session logs, decisions, patterns, and project context in .claude/memory/. Internal for session/memory workflows only.
Share bugs, ideas, or general feedback.
A skill for persisting findings across skill runs so that future skills can compare, correlate, and build on what was previously discovered.
Output type: Memory file creation and recall. This skill writes structured session files and retrieves previous session data. It does not produce analysis — it provides the memory layer that other skills draw from.
Every skill in Design System Ops produces findings. But those findings are ephemeral — they exist in the conversation, get summarised, and then vanish. The next time someone runs token-audit, it starts from zero. It cannot tell you whether things got better or worse since last quarter. It cannot cross-reference what drift-detection found last month with what component-audit finds today.
Session Memory fixes this. It creates a structured memory layer that:
This is not a database. It is a structured markdown file per session that accumulates over time. Simple, portable, and readable by both humans and AI.
This skill is a persistence layer — it saves, recalls, compares, and correlates findings. It does not run audits, produce reports, or generate recommendations on its own. If no previous session files exist and the mode is Recall, Compare, or Correlate, inform the user that there is no history to work with and suggest running an audit first. If the session memory directory does not exist, create it on Save. Do not correlate across skills unless at least two different skill sessions exist — a single-skill history is a Compare, not a Correlate.
Check for .ds-ops-config.yml in the project root:
memory:
directory: ".ds-ops/sessions" # Where session files are stored
retain_count: 12 # How many session files to keep
auto_save: true # Automatically save after every skill run
comparison_window: 3 # How many previous sessions to compare against
If no configuration exists, use these defaults:
.ds-ops/sessions/Each session file is a structured markdown document saved to the memory directory. One file per skill run.
[YYYY-MM-DD]-[skill-name]-[short-hash].md
Example: 2026-03-09-token-audit-a7f3.md
The short hash is the first 4 characters of a hash of the input parameters, ensuring uniqueness when the same skill is run twice on the same day with different inputs.
---
skill: token-audit
date: 2026-03-09
scope: "AGDS token architecture"
system: "Australian Government Design System"
component_count: 42
---
# Session: token-audit — 2026-03-09
## Summary
[2–3 sentence summary of what was found]
## Key metrics
| Metric | Value |
|---|---|
| Total tokens | 140 |
| Violations | 12 |
| Critical | 2 |
| High | 4 |
| Orphaned tokens | 45 (32%) |
## Findings
### TA-01 [Critical] — Missing semantic tier for feedback tokens
[One paragraph description]
### TA-02 [High] — Orphaned primitive tokens in legacy palette
[One paragraph description]
[... all findings ...]
## Recommendations
1. [First recommendation]
2. [Second recommendation]
## Cross-references
- Correlates with: [any related findings from prior sessions, or "None — first run"]
- Trend: [Improving / Stable / Worsening / New baseline]
Session Memory has four modes. Determine which one based on the request:
Someone ran a skill and wants the findings persisted.
Someone wants to see what was found in previous runs before running a new skill.
Someone wants to compare current findings against previous runs to see trends.
Someone wants to cross-reference findings across different skills to surface persistent patterns.
Accept findings from any skill output: copy-pasted report, file reference, or inline conversation output.
Parse the skill output and extract:
Check for prior sessions from the same skill:
## Trend section comparing key metricsCheck for cross-skill correlations:
## Cross-references sectionWrite the session file to the memory directory using the filename convention
Prune old sessions if the total exceeds retain_count. Delete the oldest files first. Never delete files from the current quarter.
Confirmation message:
Session saved: 2026-03-09-token-audit-a7f3.md
Location: .ds-ops/sessions/
Findings: 12 (2 Critical, 4 High, 4 Medium, 2 Low)
Trend: Improving — violations down from 18 to 12 since 2026-01-15
Cross-references: 2 correlations found with previous component-audit run
A skill name, a date range, or "everything."
## Session history: token-audit
| Date | Violations | Critical | Trend |
|---|---|---|---|
| 2026-03-09 | 12 | 2 | ↓ Improving |
| 2026-01-15 | 18 | 4 | ↑ Worsening |
| 2025-10-20 | 14 | 3 | Baseline |
### Most recent findings (2026-03-09)
- TA-01 [Critical] — Missing semantic tier for feedback tokens
- TA-02 [High] — Orphaned primitive tokens in legacy palette
[...]
Two specific session files, or "compare latest with previous."
## Comparison: token-audit
**Period:** 2026-01-15 → 2026-03-09 (53 days)
### Metric deltas
| Metric | Previous | Current | Delta |
|---|---|---|---|
| Total violations | 18 | 12 | ↓ 6 (−33%) |
| Critical | 4 | 2 | ↓ 2 (−50%) |
| Orphaned tokens | 52 (38%) | 45 (32%) | ↓ 7 (−6pp) |
### Finding status
| Status | Count | Findings |
|---|---|---|
| Resolved | 4 | TA-04, TA-07, TA-12, TA-15 |
| Persistent | 8 | TA-01, TA-02, TA-03, TA-05, TA-06, TA-08, TA-09, TA-11 |
| New | 2 | TA-16, TA-17 |
| Changed severity | 2 | TA-03 (Critical → High), TA-09 (High → Medium) |
### Interpretation
The system is improving. Critical violations dropped by 50%, and four findings from the previous run have been resolved. Two new findings emerged — both related to a new token category introduced since the last audit. The persistent findings (TA-01, TA-02) should be prioritised for the next sprint.
"What patterns persist across skills?" or a specific area to investigate.
## Cross-skill correlation report
**Sessions analysed:** 5 (token-audit, naming-audit, component-audit, drift-detection, system-health)
**Date range:** 2026-01-15 to 2026-03-09
### Confirmed systemic issues (4+ skills)
**Feedback component area**
- token-audit: Missing semantic tier for feedback tokens (TA-01, Critical)
- naming-audit: Inconsistent naming in alert/toast/notification components (NA-06, High)
- component-audit: No feedback component in library (CA-03, High)
- drift-detection: 4 teams building custom feedback patterns (DD-08, High)
- Impact: Every skill surfaces this gap. This is the system's highest-priority structural gap.
### Probable correlations (3 skills)
**Legacy colour palette**
- token-audit: 45 orphaned primitive tokens from legacy palette (TA-02, High)
- naming-audit: Legacy tokens use different naming convention (NA-02, Medium)
- component-audit: 3 components still reference legacy palette directly (CA-11, Medium)
### Possible correlations (2 skills)
[...]
Any skill can trigger a session save by including this instruction at the end of its output:
Save these findings to session memory.
Any skill can request previous findings by including:
Before starting, recall the most recent session for [skill-name].
The full-system-diagnostic agent should automatically save each phase's output and load all previous sessions at the start of Phase 3 (synthesis).