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.
npx agentmods add skills/agenttoolkit/altk-evolve/retentionnpx skills add AgentToolkit/altk-evolve --skill retentiongit clone --depth 1 https://github.com/AgentToolkit/altk-evolveWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00030 | $0.01041 |
| Opus 5 | $0.00015 | $0.00521 |
| Sonnet 5 | $0.00006 | $0.00208 |
| Haiku 4.5 | $0.00003 | $0.00104 |
Grade A, and why
retention 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 3d 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.
This is a copy
100% identical to evolve-lite:retention — 6 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Retention
Overview
Runs the data-retention rules configured in evolve.config.yaml against the
local .evolve/ store: private entities under .evolve/entities/ and session
transcripts under .evolve/trajectories/. Rules match by entity type plus age
(max_age_days, from file mtime) or disuse (max_unused_days, from recall
rows in .evolve/audit.log), and either flag (a non-destructive frontmatter
marker) or delete. A delete rule on trajectories with
cascade_derived: true also deletes the entities derived from those sessions
(linked by their trajectory: frontmatter).
The script is dry-run by default — it never mutates anything unless
--apply is passed.
Workflow
Step 1: Require rules
Read evolve.config.yaml. If there is no retention: block with a rules:
list, show the user this example and stop:
retention:
rules:
- name: stale-guidelines
entity_type: guideline
max_age_days: 90
action: flag
- name: old-sessions
entity_type: trajectory
max_age_days: 365
action: delete
cascade_derived: true
Each rule needs a name and at least one of max_age_days /
max_unused_days; action is flag (default) or delete. Rules are checked
top-to-bottom and the first match wins, so put narrow rules first — and put a
longer-threshold delete before a shorter-threshold flag on the same type,
or the flag shadows the delete and it never fires.
A delete rule may also set on_missing_access_signal to say what happens when
an unused match has no recorded recall (disuse measured from file mtime):
skip (default, fail-safe — spare it and report it as skipped), flag
(downgrade the delete to a non-destructive flag), or delete (delete on the
mtime fallback anyway). The safe default means an unused delete never destroys
a memory the agent simply never recorded a recall for.
Step 2: Dry run
From the project root:
python3 ${CLAUDE_PLUGIN_ROOT}/skills/evolve-lite/retention/scripts/run_retention.py
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 3d ago First seen · 101 lines · 30 tokens per session scan A e618dbacec71
retention is a skill published in the GitHub repository AgentToolkit/altk-evolve (105 stars, last pushed 8d ago), licensed Apache-2.0. It adds 30 tokens to every session and 1,041 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to evolve-lite:retention, differing in 6 lines, and is treated as a copy.
Other skills, from other repositories
agent-recall
Persistent compounding memory for AI agents. 5 default MCP tools: sessionstart, sessionend, remember, recall, check. Full surface (18 tools) available with --full flag. Two-verb model: inhale (sessionstart) and exhale (sessionend). Correction-first memory with decision trail tracking, watchfor warnings, palace rooms…
shelve
Offload a closed conversation topic (or a whole imported dialog) to the memory shelf as a Markdown episode with a validated digest. Use when a topic is finished, when context grows heavy, before compaction, or when the user asks to shelve/archive part of the conversation. M0 prompt-only version — the agent does the…
speckit-tasks
Generate an actionable, dependency-ordered tasks.md for the feature based on available design artifacts.
speckit-clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
speckit-implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.
speckit-taskstoissues
Convert existing tasks into actionable, dependency-ordered GitHub issues for the feature based on available design artifacts.