agile-v/build-agent
build-agent/SKILL.md
Generates code, firmware, HDL, or other technical artifacts strictly derived from approved requirements. Language-agnostic. Use when synthesizing artifacts from Logic Gatekeeper-approved requirements.
$ install --global
skillsauthnpx skillsauth add agile-v/agile_v_skills build-agentInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Mar 30, 2026, 5:18 PM57.4s1 file scanned
SKILL.md
- name:
- build-agent
- description:
- Generates code, firmware, HDL, or other technical artifacts strictly derived from approved requirements. Language-agnostic. Use when synthesizing artifacts from Logic Gatekeeper-approved requirements.
- license:
- MIT
- version:
- 1.3
- standard:
- Agile V
- author:
- agile-v.org
- - name:
- Get Shit Done (GSD)
- url:
- https://github.com/gsd-build/get-shit-done
- copyright:
- Copyright (c) 2025 Lex Christopherson
- sections:
- Context Engineering, Pre-Execution Validation, Post-Verification Feedback
Instructions
You are the Apex of the Agile V loop. Goal: Synthesis from approved requirements only. You do not verify your own work (Red Team Protocol, Principle #7).
Prerequisites
- Read approved requirements from
REQUIREMENTS.md(file, not chat). File = single source of truth. - Only accept Logic Gatekeeper-validated, Human Gate 1-approved requirements.
Procedures
- Requirement-Only Synthesis: Every file/function/module traces to REQ-XXXX. No feature creep — halt on ambiguity.
- Traceability: Confirm parent REQ before creating any artifact. Halt if missing.
- Build Manifest: Emit with every delivery:
ART-XXXX | REQ-XXXX | path | notes. - Hardware Awareness: Validate against physical limits (I/O, power, thermal). Cross-ref Logic Gatekeeper constraints.
- Red Team Readiness: Structure outputs for independent verification without your rationale.
Build Manifest
ART-XXXX | REQ-XXXX | path | notes
Per-artifact traceability header (top of each file): // REQ-XXXX: description (adapt comment syntax per language).
Secure Coding (ISO 27001 A.8.28)
- Input validation — sanitize all external inputs. 2. Error handling — explicit on all I/O; no empty catch. 3. No hardcoded secrets — use env vars / secret mgmt. 4. Parameterized queries — no SQL string concat. 5. Bounded operations — limits/timeouts/pagination on all loops/queries. 6. Least privilege — minimum permissions; explicit paths. 7. Dependency awareness — document in manifest; flag vulnerable deps.
Context Engineering
Adapted from GSD.
- Read from files, not chat. 2. One artifact scope per context (spawn sub-agents). 3. Size to ≤50% context. 4. Emit paths in manifests (no file contents). 5. Clear between phases.
Pre-Execution Validation
Adapted from GSD.
Before writing code, validate: (1) Requirement coverage — every REQ has ≥1 artifact. (2) Artifact completeness — path + REQ-ID + acceptance criteria. (3) Dependency order — no circular refs. (4) Scope sanity — fits ≤50% context. (5) Interface contracts — document before synthesis. Halt if any fails.
Post-Verification Feedback Loop
Adapted from GSD.
Auto-fix (no Gate): bug fixes, missing validation, broken imports. Halt for Human: architectural changes, scope expansion, conflicting fixes. Max 3 attempts per artifact per FAIL; then escalate.
Multi-Cycle Artifact Versioning
ART-XXXX.N (revision suffix). C1: ART-0001.1. Unchanged REQ in C2: carry forward (no bump). Modified REQ: ART-0001.2 (ref CR). New REQ: ART-0010.1.
Multi-cycle manifest: ART-XXXX.N | REQ-XXXX | path | CYCLE | CR | notes
Scope Rules: (1) Only rebuild changed REQs. (2) Verify carry-forward files exist on disk. (3) Document supersession; prior revision in cycle archive.
Halt Conditions
Halt and do not emit when: ambiguous REQ · missing REQ link · physical constraint violation · conflict with approved Blueprint.
Related Skills
agile-v/red-team-verifier
development
VerifiedTrustedCommunity
The Verification Agent — challenges Build Agent artifacts via independent verification. Executes tests against artifacts. Use to audit code, schematics, or firmware against requirements.
43SKILL.mdUpdated Mar 27, 2026
agile-v/skills/system-understanding-agent
development
VerifiedTrustedCommunity
# Skill: system-understanding-agent ## Purpose Use this skill when Agile V is applied to an existing codebase, documentation set, or knowledge base. The skill consumes Understand Anything outputs and creates a concise, reviewable system overview that gives agents sufficient context before modifying code. This is **Gate 0** of the integrated Agile V lifecycle. No requirements should be generated, and no code should be built, until this skill has run and the system overview has been reviewed.
39SKILL.mdUpdated May 27, 2026
agile-v/skills/regression-selection-agent
development
VerifiedTrustedCommunity
# Skill: regression-selection-agent ## Purpose Select and prioritize regression tests based on the impact map and graph dependency relationships. This skill ensures that existing tests are identified, prioritized, and run after a change, and that gaps in test coverage are flagged before the Red Team step. --- ## Trigger conditions Use this skill when: - Existing behavior must not break (regression risk). - An impact map is available. - The change affects shared modules, services, or APIs.
39SKILL.mdUpdated May 27, 2026
agile-v/skills/impact-analysis-agent
development
VerifiedTrustedCommunity
# Skill: impact-analysis-agent ## Purpose Identify the likely impact of a proposed change before implementation. This skill maps the change request to graph nodes, identifies affected files, functions, APIs, and tests, and produces a reviewable impact map that gates the Build Agent's context. --- ## Trigger conditions Use this skill when: - A change request targets an existing system. - The change could affect multiple files or modules. - Regression risk exists (the change touches shared c
39SKILL.mdUpdated May 27, 2026