Ready to supercharge your AI agents?

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

Get Started Free →

template-sia — measured AI agent skill | DecimalAI

docxology/template-sia

SkillScore91Maintained

SIA (Self-Improvement Agent) harness exemplar — Meta/Target/Feedback loops, public/private splits, fixture replay, fail-closed validation.

by docxology

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

decimalai skills pull docxology-template-sia

Get this skill — curl · raw file

curl -fsSL https://app.decimal.ai/s/docxology-template-sia/SKILL.md

Raw file: /s/docxology-template-sia/SKILL.md ↗ · machine-readable summary at /s/docxology-template-sia.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 6d ago

—-27% tokens0% turns91% of whole cases pass with skill (20/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-01	✗→✓	▲ Improved	-17%	0%
case-04	✗→✓	▲ Improved	-49%	0%
case-06	✗→✓	▲ Improved	6%	0%
case-08	✗→✓	▲ Improved	-45%	0%
case-09	✗→✓	▲ Improved	-52%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

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

docxology / template

projects/templates/template_sia/.agents/skills/template-sia/SKILL.md

View on GitHub ↗

License MITImported 8/1/2026Version v1

Skill contents

SKILL.md

template-sia

Project-scoped skill for the in-repo exemplar at projects/templates/template_sia/. Load this when working inside the project.

When to Use

Working inside the template_sia exemplar — running scripts, editing source,

or regenerating outputs.

Forking this exemplar as the starting scaffold for a new research project.
Validating that the exemplar's contracts (thin-orchestrator, layer boundaries,

no-mocks testing) still hold after changes.

Quick Reference

bash
# From the repository root
uv run pytest projects/templates/template_sia/tests --cov=projects/templates/template_sia/src --cov-fail-under=90
uv run python scripts/pipeline/stage_02_analysis.py --project templates/template_sia
uv run python scripts/pipeline/stage_03_render.py --project templates/template_sia
uv run python scripts/pipeline/stage_04_validate.py --project templates/template_sia
uv run python scripts/pipeline/stage_05_copy.py --project templates/template_sia

Pitfalls

Keep scripts thin. Business logic belongs in src/ or shared

infrastructure/, not in scripts/. src/loop.py owns the project adapter; scripts/run_sia_loop.py only parses flags and presents result paths.

No mocks. All tests must use real data, real files, and real

computation.

Outputs are disposable. Never hand-edit output/ — regenerate from

source and config.

Run from the repo root. Commands assume the template monorepo root

as working directory unless the child AGENTS.md states otherwise.

Cross-refs

Project contract: AGENTS.md
README: README.md
TODO: TODO.md
Exemplar roster: projects/AGENTS.md

Want to see if docxology/template-sia would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.