graph-handoff

星标6

分支3

更新时间2026年7月24日 00:01

Execute the HANDOFF phase of the lifecycle via the `agf` CLI — PR creation, memory capture, knowledge export, doc completeness validation

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

来源

DiegoNogueiraDev

DiegoNogueiraDev/agf

打开 GitHub 仓库查看创作者相关仓库

下载

graph-handoff

HANDOFF phase via the agf CLI (ZERO MCP). Creates PRs, captures decisions as memories, exports knowledge, finalizes docs for delivery.

When to Use

After REVIEW approves the changes
Creating a PR for the sprint's work
Capturing decisions/knowledge for future cycles
agf phase reports HANDOFF

Mandatory Flow

agf memory write → agf snapshot create → agf export → [create PR] → agf gate handoff → agf phase DEPLOY

Workflow

Step 1: Capture Decisions

agf memory write <name>

Record architectural choices, error patterns, perf insights, integration learnings (land under workflow-graph/memories/).

Step 2: Snapshot

agf snapshot create

Step 3: Export Deliverables

agf export
agf export --format mermaid

Summary + visual overview for the PR description.

Step 4: Share Knowledge

agf export also produces a cross-project knowledge package (memory + docs, quality-filtered, deduped by hash — safe to re-run).

Step 5: Prepare Commit

git status
git diff --staged
git log --oneline -5

Commit following project conventions.

Step 6: Create PR (if requested)

gh pr create --title "<title>" --body "<body>"

Body: summary (from export), tasks done (node IDs/titles), test results, breaking changes, mermaid graph.

Step 7-8: Validate Gate

agf gate handoff

Criteria: all sprint tasks done, snapshot created, knowledge exported, memories captured, PR created (if applicable), docs updated (CLAUDE.md/ADRs/API docs).

Step 9: Transition

agf phase DEPLOY

Follow the nextAction from agf phase.

Output Format

Phase: HANDOFF → DEPLOY
PR: #N (url)  Snapshot: created
Memories: N captured  Knowledge: M docs exported
Gate: handoff_ready N/100
Status: Ready for DEPLOY

Loop link → DEPLOY (graph-deploy): agf gate deploy (harness ≥ 70).

Anti-Patterns

Don't skip agf memory write — lost decisions are rediscovered expensively
Don't skip snapshots — they're the audit trail
No PR without test results in the body
Update CLAUDE.md with new conventions
Don't ignore nextAction from agf phase
Don't skip the knowledge export — it enables cross-project learning

Codex Notes

In Codex Plan Mode, plan only — don't mutate files.
During implementation, follow AGENTS.md and use apply_patch for manual edits.

Token Economy

Economia de tokens. Os levers compartilhados por todas as skills — --select, agf retrieve-command, agf exec chain, reuso antes de criação — vivem em _shared.md → Token Economy. Fonte única: um parágrafo repetido em trinta arquivos é o trigésimo primeiro que envelhece sozinho.

同仓库更多 Skills

同仓库

graph-accessibility

DiegoNogueiraDev/agf

Accessibility compliance audit using WCAG 2.2 AA standards, ARIA validation, screen reader testing, keyboard navigation, color contrast analysis, and i18n readiness

2026-07-246

graph-analyze

DiegoNogueiraDev/agf

Execute the ANALYZE phase of the lifecycle via the `agf` CLI — PRD creation, requirements, Definition of Ready (7 checks), cross-project learning

2026-07-246

graph-api-design

DiegoNogueiraDev/agf

API governance and design audit using OpenAPI/Swagger spec generation, REST maturity model, contract validation, and breaking change detection

2026-07-246

graph-architecture

DiegoNogueiraDev/agf

Architecture governance using C4 Model, ADR lifecycle, Architecture Fitness Functions, layer boundary enforcement, and drift detection

2026-07-246

graph-backlog-generation

DiegoNogueiraDev/agf

Human-in-the-loop PLANNING skill — investigates the project (graph + git + harness/gaps) and runs the whole ANALYZE→DESIGN→PLAN chain in one faceted loop to produce a COMPLETE PRD injected as graph backlog (epics, tasks, testable AC) for a separate agent to implement. Applies the project's planning methodologies — Impact Mapping + OKR per epic, JTBD, MoSCoW, WSJF/Cost-of-Delay, User Story Mapping, Example Mapping (Rules/Examples → Given-When-Then AC), SPIDR splitting, INVEST, Definition of Ready, Risk Matrix; the full catalogue lives in the skill body. Stops for the human after each complete PRD and iterates the next cycle from the project's own findings (dogfood). Does NOT implement. Triggers — graph-backlog-generation, gerar backlog, criar PRD, planejar feature, detalhar épico, novo ciclo, "plan the next thing", "what should we build next".

2026-07-246

graph-bug-hunter

DiegoNogueiraDev/agf

Automated bug discovery through static analysis, LSP diagnostics, pattern detection, regression hotspot analysis, and error catalog mining

2026-07-246

name	graph-handoff
description	Execute the HANDOFF phase of the lifecycle via the `agf` CLI — PR creation, memory capture, knowledge export, doc completeness validation
triggers	["graph-handoff"]
version	2.0.0
author	Diego Nogueira
date	"2026-04-04T00:00:00.000Z"
toolchain	["agf memory write","agf snapshot create","agf export","agf gate","agf phase"]

graph-handoff

HANDOFF phase via the agf CLI (ZERO MCP). Creates PRs, captures decisions as memories, exports knowledge, finalizes docs for delivery.

When to Use

After REVIEW approves the changes
Creating a PR for the sprint's work
Capturing decisions/knowledge for future cycles
agf phase reports HANDOFF

Mandatory Flow

agf memory write → agf snapshot create → agf export → [create PR] → agf gate handoff → agf phase DEPLOY

Workflow

Step 1: Capture Decisions

agf memory write <name>

Record architectural choices, error patterns, perf insights, integration learnings (land under workflow-graph/memories/).

Step 2: Snapshot

agf snapshot create

Step 3: Export Deliverables

agf export
agf export --format mermaid

Summary + visual overview for the PR description.

Step 4: Share Knowledge

agf export also produces a cross-project knowledge package (memory + docs, quality-filtered, deduped by hash — safe to re-run).

Step 5: Prepare Commit

git status
git diff --staged
git log --oneline -5

Commit following project conventions.

Step 6: Create PR (if requested)

gh pr create --title "<title>" --body "<body>"

Body: summary (from export), tasks done (node IDs/titles), test results, breaking changes, mermaid graph.

Step 7-8: Validate Gate

agf gate handoff

Criteria: all sprint tasks done, snapshot created, knowledge exported, memories captured, PR created (if applicable), docs updated (CLAUDE.md/ADRs/API docs).

Step 9: Transition

agf phase DEPLOY

Follow the nextAction from agf phase.

Output Format

Phase: HANDOFF → DEPLOY
PR: #N (url)  Snapshot: created
Memories: N captured  Knowledge: M docs exported
Gate: handoff_ready N/100
Status: Ready for DEPLOY

Loop link → DEPLOY (graph-deploy): agf gate deploy (harness ≥ 70).

Anti-Patterns

Don't skip agf memory write — lost decisions are rediscovered expensively
Don't skip snapshots — they're the audit trail
No PR without test results in the body
Update CLAUDE.md with new conventions
Don't ignore nextAction from agf phase
Don't skip the knowledge export — it enables cross-project learning

Codex Notes

In Codex Plan Mode, plan only — don't mutate files.
During implementation, follow AGENTS.md and use apply_patch for manual edits.

Token Economy

Economia de tokens. Os levers compartilhados por todas as skills — --select, agf retrieve-command, agf exec chain, reuso antes de criação — vivem em _shared.md → Token Economy. Fonte única: um parágrafo repetido em trinta arquivos é o trigésimo primeiro que envelhece sozinho.