Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

SkillOpt-Sleep: usage-driven self-evolution for the dsh agent · microsoft | DecimalAI

agent toolingby microsoft

SkillOpt-Sleep: usage-driven self-evolution for the dsh agent

microsoft/skillopt-sleep

Source microsoft/skillopt ↗License MIT

Reference-only OpenClaw adaptation of SkillOpt-Sleep. Use it to study or port the contributed DeepSeek wrapper, not as a ready-to-run installation.

SkillScore80Maintained

Automatic

AI Engineer pack

Last verified 8/23/2026

What it changes

Passed 18 of 23 test cases with the skill, and 9 of 23 without it.

Without skill40%

With skill80%

Checks it passed that the model alone missed

Includes the --assert-improves flag
Notes that skillopt_dry_run stages nothing
Invokes skillopt_unschedule for the project directory

and 10 more in the full report

10 fixed10 passed either way5 still failing0 broke

Pass-rate lift: +40 pp
Tokens: +25%
Agent turns: 0%
Response time: -49%

Model gemini-3.6-flashMeasured 8/23/2026v1020 of 25 cases passed with skill · 23 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/microsoft-skillopt-sleep/SKILL.md https://app.decimal.ai/s/microsoft-skillopt-sleep/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add microsoft-skillopt-sleep, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/microsoft-skillopt-sleep/ and pins it in .decimal/skills.lock

decimalai skills add microsoft-skillopt-sleep

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/microsoft-skillopt-sleep/SKILL.md https://app.decimal.ai/s/microsoft-skillopt-sleep/SKILL.md

Raw file: /s/microsoft-skillopt-sleep/SKILL.md ↗ · machine-readable summary at /s/microsoft-skillopt-sleep.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 10

Prompt

We want to stop the automated nightly evolution schedule for project `/home/dev/service`. Which tool command should be executed?

What was checked, without and with the skill
Checked	Without skill	With skill
Invokes skillopt_unschedule for the project directory	Missed	Passed

case-16 · gemini-3.6-flash · 8/23/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md116 linessingle file

SkillOpt-Sleep: usage-driven self-evolution for the dsh agent

SkillOpt-Sleep is Microsoft's SkillOpt deployment-time companion engine: it reviews your past sessions (harvest), mines recurring tasks (mine), replays them through a selected backend (replay), and consolidates what it learns into skill documents behind a held-out validation gate (consolidate).

This skill drives the engine through the 7 skillopt_* tools exposed by the dsh-skillopt plugin. The default mock backend makes no model calls, which is useful for verifying the plumbing; a real backend consumes your API budget.

When to use

"make my agent better the more I use it" / "learn my preferences across sessions"
a one-off offline self-evolution / sleep / dream run (immediate or scheduled)
review past sessions/trajectories and distill recurring tasks
consolidate feedback into AGENTS.md / SKILL.md / managed skills
schedule (cron) the cycle, or adopt a staged proposal

The cycle (six stages)

Harvest — read-only scan of supported local session records → digests
Mine — digests → recurring task records (intent + outcome labels + checkable refs)
Replay — re-run tasks under the current skill+memory with the selected backend → (hard, soft) scores
Consolidate — reflect on failures → propose bounded edits → validation gate on a held-out slice (default: accept only on strict improvement)
Stage — write accepted proposals to <project>/.skillopt-sleep/staging/<timestamp>/. Live files are unchanged. A rejected run still has a report but no proposal files.
Adopt — explicit (or operator-configured --auto-adopt) copies staged files over live ones, backing up first.

Driving it

Prefer the tools over hand-editing files:

Tool	Behavior
`skillopt_status`	state, engine availability, latest staged proposal & report
`skillopt_dry_run`	full preview (harvest+mine+replay), stages nothing
`skillopt_run`	full cycle, stages a proposal (live files unchanged by default)
`skillopt_adopt`	apply latest staged proposal (with backup) — the live-change boundary
`skillopt_harvest`	read-only show/export of mined tasks

Track record

Benchmark

18 / 23

test cases passed, vs 9 / 23 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 8/23/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

microsoft / skillopt

plugins/openclaw/SKILL.md

View on GitHub ↗

License MITImported 8/23/2026Version v10

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

SkillOpt-Sleep: usage-driven self-evolution for the dsh agent

When to use

The cycle (six stages)

Driving it

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills