timgranlundmarsden

timgranlundmarsden/claude-agent-flow

Hook Claude Code

Runs when a session starts, executing session-start.sh. From timgranlundmarsden/claude-agent-flow.

timgranlundmarsden/claude-agent-flow

Hook Claude Code

Runs after a tool call finishes for Bash, Skill and mcp__*backlog* tool calls, running an inline shell check (3 commands). From timgranlundmarsden/claude-agent-flow.

timgranlundmarsden/claude-agent-flow

MCP server Claude CodeCodexCursor

MCP server "backlog" as configured in timgranlundmarsden/claude-agent-flow. Launched with backlog mcp start.

timgranlundmarsden/claude-agent-flow

Plugin Claude Code

Multi-agent development pipeline for Claude Code — 12 specialist agents, 13 slash commands, 18 skills, adversarial critic loop, and automated sync distribution. Drop-in plugin that adds /plan, /build, /review workflows with explorer, architect, builder, critic, tester, and reviewer agents.

timgranlundmarsden/claude-agent-flow

Settings file Claude Code

Agent settings declaring 2 hook events (SessionStart, PostToolUse) and 64 allowed tools.

timgranlundmarsden/claude-agent-flow

Instructions file

Claude Code instructions for timgranlundmarsden/claude-agent-flow, covering project: agent-flow, backlog management, git workflow — required before every push, main branch protection and failure-mode contract.

timgranlundmarsden/claude-agent-flow

Agent

Adversarial code critic. Tries to break code with edge cases and failure scenarios. Returns FAIL/PASS. Used in /build and /review loops.

timgranlundmarsden/claude-agent-flow

Agent

Read-only codebase navigator. Maps files, traces dependencies, surfaces existing patterns. Cheap — use constantly before implementation tasks.

timgranlundmarsden/claude-agent-flow

Command

Skills: agent-flow-init-check.

timgranlundmarsden/claude-agent-flow

Command

Display a live backlog status dashboard — active tasks table with PR state and branch-to-test. Optional flag: --qa also generates a QA test guide for all Ready-for-Review tasks.

timgranlundmarsden/claude-agent-flow

Command

Build a feature using the adversarial review loop. Default tier: parent works end-to-end, zero subagent spawns. --strict tier adds critic + external review.

timgranlundmarsden/claude-agent-flow

Command

Check the current branch's PR for external review feedback, fix errors and warnings, run a final external review, and push.

timgranlundmarsden/claude-agent-flow

Command

Report the health of the agent-flow installation.

timgranlundmarsden/claude-agent-flow

Command

Quick read-only exploration of the codebase. Invokes explorer agent to map files, dependencies, and patterns relevant to a topic. No code changes. Use before starting any significant task to understand the landscape.

timgranlundmarsden/claude-agent-flow

Command

Run a single external code review via an external LLM API on the full branch diff vs main. Standalone usage — no aggregation with internal review.

timgranlundmarsden/claude-agent-flow

Command

Show comprehensive help guide for the agent team system.

timgranlundmarsden/claude-agent-flow

Command

Initialize or update the agent-flow system in this repository.

timgranlundmarsden/claude-agent-flow

Command

Guide a collaborative planning session. Conducts Socratic dialogue, spawns explorer on large repos and researcher on new deps; saves plan + backlog task.

timgranlundmarsden/claude-agent-flow

Command

Guided rebase workflow. Stashes uncommitted changes, fetches latest from origin, rebases onto the remote branch (or a specified branch), resolves conflicts if possible, and pops the stash on success.

timgranlundmarsden/claude-agent-flow

Command

Run the adversarial critic against existing code. Default: parent reviews inline, zero subagent spawns. --strict: critic subagent + external review.

timgranlundmarsden/claude-agent-flow

Command

View, create, or update TECHSTACK.md for the current project. Runs the Technology Discovery Protocol to detect the project's stack from source files, package manifests, and CI/CD config. Safe to run at any time — existing user-added content is always preserved. Use --force to bypass the 72-hour freshness check and…

timgranlundmarsden/claude-agent-flow

Command

Analyse Claude Code session token costs inline.

timgranlundmarsden/claude-agent-flow

Skill Claude CodeCodex

This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development…

timgranlundmarsden/claude-agent-flow

Skill Claude CodeCodex

Guard skill that halts command execution if the agent-flow system is not initialized. Checks for .claude-agent-flow/sync-state.json (installed repos) or .claude-agent-flow/publish-plugin-manifest.yml (master source repo) and stops with a clear message to run /install if neither is present.