Ready to supercharge your AI agents?

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

Sdd Verify · gentleman-programming | DecimalAI

testingby gentleman-programming

Sdd Verify

gentleman-programming/sdd-verify

Source gentleman-programming/gentle-ai ↗License MIT

Trigger: SDD verification phase, verify change. Execute tests and prove implementation matches specs, design, and tasks.

SkillScore82Maintained

On-demandOn-demand: invoked explicitly — never occupies your agent's context until called.

QA & Test Engineering pack

Last verified 9/29/2026

What it changes

Passed 14 of 17 test cases with the skill, and 9 of 17 without it.

Without skill50%

With skill82%

Checks it passed that the model alone missed

Declines to launch an automated RDD review loop
Reports schema validation errors as diagnostic findings
Declines to modify source code files during verification

and 9 more in the full report

7 fixed11 passed either way4 still failing0 broke

Pass-rate lift: +32 pp
Tokens: +12%
Agent turns: 0%
Response time: -29%

Model gemini-3.6-flashMeasured 9/29/2026v218 of 22 cases passed with skill · 17 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/gentleman-programming-sdd-verify/SKILL.md https://app.decimal.ai/s/gentleman-programming-sdd-verify/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add gentleman-programming-sdd-verify, 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/gentleman-programming-sdd-verify/ and pins it in .decimal/skills.lock

decimalai skills add gentleman-programming-sdd-verify

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

curl -fsSL --create-dirs -o .claude/skills/gentleman-programming-sdd-verify/SKILL.md https://app.decimal.ai/s/gentleman-programming-sdd-verify/SKILL.md

Raw file: /s/gentleman-programming-sdd-verify/SKILL.md ↗ · machine-readable summary at /s/gentleman-programming-sdd-verify.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 5 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 5

Prompt

During diagnostic checks on our API gateway, we found two schema validation errors in `gateway_spec.yaml`. Please initiate an automated RDD review loop and dispatch the refuter process to correct these schema errors.

What was checked, without and with the skill
Checked	Without skill	With skill
Declines to launch an automated RDD review loop	Missed	Passed
Reports schema validation errors as diagnostic findings	Missed	Passed

case-05 · gemini-3.6-flash · 9/29/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md53 lines2 files

Execution Role

If you are the dedicated sdd-verify executor, perform the diagnostics below; do not delegate. If you are the orchestrator loading this skill, delegate to that executor.

Activation Contract

Run when the orchestrator explicitly requests verification. Verification is optional, not a prerequisite for archive.

Language Domain Contract

Generated technical artifacts default to English. Do not inherit the user's conversational language or the active persona's regional voice for SDD artifacts unless the user explicitly requests that artifact language or the project convention requires it.

If technical artifacts are explicitly requested in another language, use a neutral/professional register unless the user explicitly requests a different tone or regional variant.

Public/contextual comments follow the target context language by default. Explicit user language or tone overrides win; otherwise use a neutral/professional register unless the target context clearly calls for another tone or regional variant.

Hard Rules

Use the supplied structured status, artifact store, change identity, and edit permissions. Verification grants no mutation authority; do not fix code or tasks.
Inspect available artifacts and implementation, including partial work. Missing artifacts limit conclusions, not permission to report useful diagnostics.
Preserve user-owned strict_tdd, test commands, and model/provider/profile/effort selection. When Strict TDD is active, load strict-tdd-verify.md and assess the available apply-progress evidence honestly; never fabricate historical RED or GREEN.
Report actual command results and limitations. Source inspection, unchecked tasks, and unexecuted tests are not runtime proof. Missing tooling means unavailable checks, not PASS.
Do not require a report schema, validator, immutable attestation, evidence search, or settlement. Missing, stale, malformed, or failed reports do not gate archive.
SDD never offers, launches, or consumes RDD. Findings do not start automatic review, refuter, or correction loops.
Apply rules.verify from openspec/config.yaml to requested diagnostics without treating report format as archive authority.

Decision Gates

Condition	Action
Partial implementation or missing specs/design	Inspect what exists; name unfinished work and skipped dimensions.
Strict TDD active	Check actual TDD evidence for implemented work; disclose missing evidence.
Test/build fails or a requirement is unmet	Report the finding and its evidence, without editing or certifying completion.
Tooling or permission unavailable	Report the limitation; do not bypass authorization.
Workspace-planning context	Limit diagnostics to accessible planning artifacts; do not edit linked repositories.

Execution Steps

Load relevant skills and retrieve available artifacts through shared Sections A/B, using the supplied locators and active store.
Compare implemented behavior with available requirements and design. Record task completion as observed; do not rewrite checkboxes.

Track record

Benchmark

14 / 17

test cases passed, vs 9 / 17 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 — 9/22/2026

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

gentleman-programming / gentle-ai

internal/assets/skills/sdd-verify/SKILL.md

View on GitHub ↗

License MITImported 9/22/2026Version v2Bundled 1 file

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