tekkiiiii/the-agency/task-handoff

Claude Code and agent skill by tekkiiiii/the-agency

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when an agent finishes work that unblocks another agent, or when defining ownership in a multi-agent pipeline. Agent A writes outcome and next steps to the task store, creates Agent B's task and marks its own done; the store is the only handoff channel, not chat context.

Install

npx vskill@latest install tekkiiiii/the-agency/task-handoff

SKILL.md

The same SKILL.md was published earlier as tekkiiiii/the-agency/task-handoff.

Tier-A structured agent handoff — no new infrastructure, just discipline with the task store. The task store is the coordination layer; conversation context is not.

Protocol

When Agent A completes work that unblocks Agent B:

  1. Agent A writes its outcome to the task store — status, findings, any blockers, next steps
  2. Agent A creates Agent B's task in the task store with assigned_agent set
  3. Agent A marks its own task `done` and records the handoff ID in notes
  4. Agent B reads from the task store to find its assignment — not from memory or conversation

Example: Feature implementation → QA review

# Agent A (dev): marks feature done, creates QA task
ts-create "myproject" "qa-review-api" "QA the REST API implementation" "phase1" "high"
# -> returns downstream_id
# Agent A sets blocked_by to empty (upstream gates were cleared)
sqlite3 ~/.claude/task-store.db "UPDATE tasks SET blocked_by='[]' WHERE id='$downstream_id'"
# Agent A marks own task done
ts-status "$upstream_id" "done"
# Agent B (QA): reads task store for its assignment
ts-project "myproject" | grep "qa-review-api"
# -> finds task, starts work
# After QA, Agent B writes gate verdict
ts-gate "$feature_task_id" "passed" "qa-agent"

Key Rules

  • Task store is the source of truth — not conversation, not memory
  • `blocked` status is a hard gate — never advance to in_progress while blocked
  • Handoffs are explicit — always write to task store before spawning downstream
  • Failure propagates — if Agent B fails after retries, escalate to PD/council

Anti-patterns

  • Spawning an agent without checking the task store first
  • Storing state in conversation context (it gets lost between sessions)
  • Advancing a blocked task to in_progress
  • Skipping gate verification for "quick" downstream tasks

Relationship to Tier-B (future)

Tier-B adds a file-based FIFO queue (~/.claude/state/handoff-queue/) so agents can poll asynchronously. Tier A is the prerequisite — implement this first to establish the discipline.

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
3
7-day trend score
7

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 5 Oct 2026, a month ago.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/tekkiiiii/the-agency/task-handoff/badge)](https://verified-skill.com/skills/tekkiiiii/the-agency/task-handoff)

Something wrong with this skill or its page? Tell us.