Ready to supercharge your AI agents?

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

Get Started Free →

Requirements Authoring Flow · griddynamics | DecimalAI

documentationby griddynamics

Requirements Authoring Flow

griddynamics/requirements-authoring-flow

Source griddynamics/rosetta ↗License Apache-2.0

Workflow for authoring requirements and specifications: drafting, review, validation, etc.

SkillScore68Maintained

Automatic

Product Management pack

Last verified 8/27/2026

What it changes

Passed 15 of 22 test cases with the skill, and 5 of 22 without it.

Without skill23%

With skill68%

Checks it passed that the model alone missed

The response refuses to proceed on the current model class.
The response mandates running conflict checks and gap checks.
The response keeps unresolved requirement units marked as Draft.

and 13 more in the full report

11 fixed4 passed either way6 still failing1 broke

Pass-rate lift: +45 pp
Tokens: +23%
Agent turns: 0%
Response time: -45%

Model gemini-3.6-flashMeasured 8/27/2026v615 of 22 cases passed with skill

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/griddynamics-requirements-authoring-flow/SKILL.md https://app.decimal.ai/s/griddynamics-requirements-authoring-flow/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add griddynamics-requirements-authoring-flow, 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/griddynamics-requirements-authoring-flow/ and pins it in .decimal/skills.lock

decimalai skills add griddynamics-requirements-authoring-flow

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

curl -fsSL --create-dirs -o .claude/skills/griddynamics-requirements-authoring-flow/SKILL.md https://app.decimal.ai/s/griddynamics-requirements-authoring-flow/SKILL.md

Raw file: /s/griddynamics-requirements-authoring-flow/SKILL.md ↗ · machine-readable summary at /s/griddynamics-requirements-authoring-flow.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 11 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 11

Prompt

We are running Phase 2 (Research) to gather industry compliance standards for HIPAA-compliant medical record storage requirements. Which subagent role and required model class must be assigned to this phase?

What was checked, without and with the skill
Checked	Without skill	With skill
The response assigns Phase 2 research to the requirements-engineer subagent.	Missed	Passed
The response specifies the gpt-5.6-sol-high model for the research subagent.	Missed	Passed

case-21 · gemini-3.6-flash · 8/27/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md167 linessingle file

<description_and_purpose> Prevents premature drafting by enforcing HITL gates where every <req> unit receives explicit user approval before proceeding. Each phase produces traceable artifacts (Final Requirements Set, Validation Pack, Traceability Matrix). Input: user request for new requirements, edits, review, refactor, or validation; USE SKILL requirements-authoring and PROACTIVELY REQUIRE its use in all phases and subagents. </description_and_purpose>

<workflow_phases>

All Rosetta prep steps MUST be FULLY completed
USE SKILL load-project-context, hitl
MUST ALWAYS use todo tasks ledger, ASAP. Phases are sequential. Independent tasks can run in parallel.
No rush, take your time, MUST FOLLOW WORKFLOW ENTIRELY, no skipping
Every phase MUST update requirements-authoring-flow-state.md in FEATURE TEMP with: phase name, status, artifact produced, and open questions.
Orchestrator and subagents MUST USE SKILL requirements-authoring.
If task is to reverse engineer orchestrator MUST USE SKILL reverse-engineering.
Keep requirement identifiers in code comments only, must not be user facing.
If /goal is set repeat phases 5-6 until goal is met, then continue with the rest of phases.
This workflow MUST be used with Fable, Opus, GPT-5.5+ class models => IF NOT - DEMAND USER TO SWITCH MODEL

IMPORTANT! If the task is to reverse engineer requirements, spawn MULTIPLE subagents with each handling one unit of analysis (one screen, one page, one controller, one endpoint, etc) to effectively prevent hallucinations by narrow scoping for phases intent_capture, outline, draft, validate.

Artifact: Discovery Summary (context, existing requirements, constraints, affected files). Done when: scope boundaries and relevant requirement files are identified.

Complete all Rosetta prep steps
Detect environment and project structure
Read existing requirements, glossary, assumptions, constraints
Identify requirement areas (FR, NFR, interfaces, data, traceability)
Record assumptions and unknowns
Required skills: requirements-authoring
Recommended skills: reverse-engineering

Track record

Benchmark

15 / 22

test cases passed, vs 5 / 22 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 — 8/26/2026

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

griddynamics / rosetta

plugins/core-antigravity/skills/requirements-authoring-flow/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/26/2026Version v6

Other skills for the same job

Published skills whose descriptions closely match this one, with their headline benchmark lift.

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