–

product-capability

by affaan-m/ECC · Productivity & Automation

Turns PRDs, roadmap items, and product discussions into implementation-ready capability plans that surface constraints, invariants, and open decisions before multi-service engineering work begins.

Lift · GLM-5.3-flash+3ppdirectional · all models →

Trustsecurity · quality · license · upkeep

Scan Saferisk 0/100

LicenseMITserved via MCP

GLM-5.3-flash Claude Haiku 5.5 GPT-6 Luna

Pass-rate Lift−6pp95% CI [−19, +8] · CI includes zero

full tier · not verifiedneutral sandbox

−100pp+100pp

Pass rate with the skill minus without, same tasks, same model, same container.

cost of the lift−$0.0362 per task · −5,827,211 input tokens · aux $0.026 (generated media)agent cost with the skill minus without, per task, this model — excludes SkillzBench's own judging.

baseline

68%

with skill

62%

forced

68%

Content Lift forced arm: +1pp [−13, +13]
Trigger rate: 100%
Trials: 110 · 8 tasks
Model: Claude Haiku 5.5
Measured: 2026-10-20

Content Lift = the skill's instructions injected into the system prompt, so it measures what the content does once loaded, apart from whether the agent chooses to load it. Method →

Embed this score · live badge for your README

The badge shows the real benchmark number and updates itself when we re-bench — including the honest states (“did not trigger”, “not benched yet”). All badge options →

Markdown

[![SkillzBench](https://skillzbench.com/api/badge/product-capability.svg)](https://skillzbench.com/skills/product-capability)

+ cost/task

[![SkillzBench](https://skillzbench.com/api/badge/product-capability.svg?cost=1)](https://skillzbench.com/skills/product-capability)

HTML

<a href="https://skillzbench.com/skills/product-capability"><img src="https://skillzbench.com/api/badge/product-capability.svg" alt="SkillzBench benchmark" height="20"></a>

Real outputs, side by side

Same public task, same model, same container. Left: without the skill. Right: with it. Drag the handle. These are the grader's own renders that the blind judge scored; the Scrolled views are captured by a scripted scroll pass, so scroll-driven work (reveals, pins, parallax, progress) is visible mid-flight. Held-out task outputs are never shown.

public task · productivity-automation-149tsw6r
Turn the PRD at /work/inputs/prd_audit_trail.md into an implementation-ready capability plan for the Audit Trail capability, taking the consumer notes in /work/inputs/audit_consumers.md as binding.…

without skillwith skill

without skill · score 83/100 · pass with skill · score 76/100 · pass · skill loadedClaude Haiku 5.5 · median trial per arm · blind-judged

Limitations

Results are for the SkillzBench reference agent on Claude Haiku 5.5. Switch models above — a skill can help one model and not another.
Screen tier (1 trial per arm) is directional. Verified needs a full-tier run (≥ 8 tasks, ≥ 3 trials per arm) with the CI excluding zero.
Task coverage: 8 generated tasks for this skill's category; a skill can shine elsewhere.