self-improve

A command that displays statistics and recent activity from Workshop memory, a local store of saved project decisions, notes, standards, and preferences.

In plain words
What is it for?
Viewing memory counts, listing recent entries, or filtering saved memory by a domain such as iOS.
Why use it?
It makes it easier to see what project knowledge has been saved and inspect recent entries without opening the database directly.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/adilkalam/orca/self-improve
Clone the repo
git clone --depth 1 https://github.com/adilkalam/orca
Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 450 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00007 $0.00450
Opus 5 $0.00003 $0.00225
Sonnet 5 $0.00001 $0.00090
Haiku 4.5 $0.00001 $0.00045

Measured 2d ago against content hash 51eb278a71b3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

self-improve scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/self-improve.md · 78 lines

What it actually says

/self-improve - Workshop Memory Stats (OS 7.1)

Display statistics about the Workshop persistent memory system.


Usage

/self-improve                    # Show memory stats
/self-improve --recent           # Show recent entries
/self-improve --domain ios       # Filter by domain

Show Stats

Display counts of Workshop memory entries:

# Typed counts read directly from workshop.db (accurate).
# NOTE: save_standard persists as type 'gotcha'; there is no separate 'standard' type.
DB=.claude/memory/workshop.db
echo "Decisions:   $(sqlite3 "$DB" "SELECT COUNT(*) FROM entries WHERE type='decision';" 2>/dev/null)"
echo "Standards:   $(sqlite3 "$DB" "SELECT COUNT(*) FROM entries WHERE type='gotcha';" 2>/dev/null)"
echo "Notes:       $(sqlite3 "$DB" "SELECT COUNT(*) FROM entries WHERE type='note';" 2>/dev/null)"
echo "Preferences: $(sqlite3 "$DB" "SELECT COUNT(*) FROM entries WHERE type='preference';" 2>/dev/null)"

Output format:

## Workshop Memory Stats

| Type | Count |
|------|-------|
| Decisions | 42 |
| Standards | 18 |
| Task History | 156 |

Recent Activity

If --recent flag is set:

workshop --workspace .claude/memory recent

Domain Filter

If --domain flag is set:

workshop --workspace .claude/memory search "$DOMAIN" --limit 20

See Also

  • /project-memory - the command that wraps the Workshop CLI (status/why/decide/gotcha/search)
  • /audit - Creates standards from audit findings
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 78 lines · 7 tokens per session scan A 51eb278a71b3

Subscribe to this mod's changes

self-improve is a command published in the GitHub repository adilkalam/orca (2 stars, last pushed 1mo ago), licensed MIT. It adds 7 tokens to every session and 450 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.