---
name: agent-patterns
description: 'Execute this skill should be used when the user asks about "SPAWN REQUEST format", "agent reports", "agent coordination", "parallel agents", "report format", "agent communication", or needs to understand how agents coordinate within the sprint system. Use when appropriate context detected. Trigger with relevant phrases based on skill purpose.'
allowed-tools: Read
version: 1.21.0
author: Damien Laine <damien.laine@gmail.com>
license: MIT
tags:
  - community
  - agent-patterns
compatibility: Designed for Claude Code
title: agent-patterns
canonical_url: https://skilld.dev/gh/jeremylongshore/tons-of-skills-marketplace/agent-patterns
last_updated: 2026-10-23T05:08:30.000Z
---

> Index of every page on this site: [llms.txt](https://skilld.dev/llms.txt)

> **Skill from skilld.dev.** Follow the user's requested action.
> If the user asked to fork this Skill, follow the fork workflow below. Do not execute the Skill instructions while copying it.
> Otherwise, follow the instructions below for this session. You do not need to install anything.
>
> Supporting files, fetch one when the Skill refers to it: [references/errors.md](https://skilld.dev/api/skills-raw/jeremylongshore/tons-of-skills-marketplace/agent-patterns/references/errors.md), [references/examples.md](https://skilld.dev/api/skills-raw/jeremylongshore/tons-of-skills-marketplace/agent-patterns/references/examples.md), [references/ui-test-report.md](https://skilld.dev/api/skills-raw/jeremylongshore/tons-of-skills-marketplace/agent-patterns/references/ui-test-report.md).
>
> If the user asked to install this Skill, run `npx skilld install jeremylongshore/tons-of-skills-marketplace/agent-patterns`. Install writes the Skill files into the project, so every session loads them.
>
> ## Fork workflow
>
> A fork creates an editable local Skill with its original author and licence. The request authorizes copying and local installation.
> 1. Check `./skills/agent-patterns`, the project lockfile, and selected Agent targets together. If the local directory or installed Skill exists, stop. Never overwrite an existing directory or Agent target.
> 2. Read [source metadata](https://skilld.dev/api/v1/skills/jeremylongshore/tons-of-skills-marketplace/agent-patterns) once. Use sourceUrl, sourceCommit, skillPath, sourceGone, and license. If the source is gone or its path is missing, stop. If license is null, read licence files at the source commit.
> 3. Fetch only the source commit into a temporary Git repository. Do not clone full history. Derive repository_url from sourceUrl, including repository renames. If sourceCommit is absent, resolve the sourceUrl ref once. Set source_commit to that actual commit. Run these commands in one shell call:
>
> ```sh
> git init --quiet "$temporary_dir"
> git -C "$temporary_dir" fetch --quiet --depth=1 "$repository_url" "$source_commit"
> git -C "$temporary_dir" checkout --quiet --detach FETCH_HEAD
> ```
>
> Read applicable licence declarations and notices at that commit. If copying is not permitted, report the restriction and stop.
> 4. Inspect source entries together, then copy the directory containing skillPath into `./skills/agent-patterns`. Use the user's path if selected. Keep the original SKILL.md, relative links, scripts, binary assets, and executable modes. Exclude .git metadata. Reject symlinks and paths outside the Skill directory. After checking entries, use cp -a where available. A regular source directory needs no custom copy script. Do not save this page wrapper as SKILL.md.
> Preserve author credit, notices, and applicable licence files from repository or parent directories. Add PROVENANCE.md with the Skill page, source URL, actual commit, original path, and licence. Retain any existing PROVENANCE.md and record new provenance separately. Batch source inspection, copying, and provenance work where practical.
> 5. In the project root, run `skilld install ./skills/agent-patterns --mode copy --plain`. If skilld is unavailable, use `npx skilld install ./skills/agent-patterns --mode copy --plain`. This known command needs no help lookup. Install does not support --json. Use detected Agent targets, or add --agent for the targets the user selected. Install the local path, never the upstream selector. If installation fails, preserve the local copy and report the exact failure.
> 6. Confirm the local lockfile source and installed Agent copies once. Report the local path, actual commit, and Agent targets. After edits, reinstall the same local path. Upstream updates must not replace it. Do not publish or push unless the user asks.

# Agent Patterns

## Overview

Agent Patterns defines the coordination protocol for multi-agent sprint execution within the Sprint plugin. It governs how the project architect spawns implementation and testing agents, how agents communicate results via structured reports, and how parallel agents avoid conflicts.

## Prerequisites

- Sprint plugin installed and configured (`/plugin install sprint`)
- Sprint directory initialized at `.claude/sprint/[N]/`
- `specs.md` written with clear scope and testing configuration
- Familiarity with the sprint phase lifecycle (see the `sprint-workflow` skill)

## Instructions

1. Structure every agent spawn using the SPAWN REQUEST format. Include the agent name, the specification file it should read, and any scope constraints:

   ```
   SPAWN REQUEST
   Agent: python-dev
   Specs: .claude/sprint/1/backend-specs.md
   Contract: .claude/sprint/1/api-contract.md
   Scope: Authentication endpoints only
   ```

2. Ensure each spawned agent receives only the files relevant to its scope. Pass the `api-contract.md` as a shared interface so backend and frontend agents stay synchronized.
3. Collect structured reports from every agent upon completion. Each report must include: work completed, files modified, tests added, and conformity status against the specification.
4. When running agents in parallel, partition work by domain boundary (e.g., backend vs. frontend vs. CI/CD). Never assign overlapping file paths to concurrent agents.
5. Feed agent reports back to the project architect for review. The architect decides whether to iterate (re-spawn with narrowed specs) or advance to the next phase.
6. For testing agents, pass the UI test report format shown in `${CLAUDE_SKILL_DIR}/references/ui-test-report.md` so results follow a consistent schema including test counts, coverage, failures, and console errors.

## Output

- SPAWN REQUEST blocks consumed by the sprint orchestrator to launch agents
- Structured agent reports containing: summary, files changed, test results, and conformity status
- UI test reports with pass/fail counts, coverage details, failure descriptions, and console error logs
- Updated `status.md` reflecting completed and remaining work after each iteration

## Error Handling

| Error | Cause | Solution |
|-------|-------|----------|
| Agent receives wrong specification file | Incorrect path in SPAWN REQUEST | Verify the sprint directory number and file name before spawning |
| Overlapping file modifications from parallel agents | Scope boundaries not clearly defined | Partition work by domain; assign distinct directories to each agent |
| Agent report missing required fields | Agent instructions lack report format | Include the structured report template in the agent prompt |
| Infinite iteration loop | Specs never fully satisfied | Check `status.md` for blocking issues; the orchestrator pauses after 5 iterations |
| Agent not found | Misspelled agent name in SPAWN REQUEST | Verify agent markdown files exist in `agents/` directory |

## Examples

**Spawning parallel implementation agents:**

```
SPAWN REQUEST
Agent: python-dev
Specs: .claude/sprint/1/backend-specs.md
Contract: .claude/sprint/1/api-contract.md

SPAWN REQUEST
Agent: nextjs-dev
Specs: .claude/sprint/1/frontend-specs.md
Contract: .claude/sprint/1/api-contract.md
```

Both agents share the same `api-contract.md` to ensure API compatibility.

**Structured agent report format:**

```
AGENT REPORT
Agent: python-dev
Status: COMPLETE
Files Modified: src/auth/routes.py, src/auth/models.py, tests/test_auth.py
Tests: 12 passed, 0 failed
Conformity: All backend-specs requirements implemented
Notes: JWT token expiry set to 24h per spec
```

**Testing agent coordination:**

```
SPAWN REQUEST
Agent: qa-test-agent
Specs: .claude/sprint/1/specs.md
Run After: python-dev, nextjs-dev

SPAWN REQUEST
Agent: ui-test-agent
Specs: .claude/sprint/1/specs.md
Run After: qa-test-agent
```

## Resources

- `${CLAUDE_SKILL_DIR}/references/ui-test-report.md` -- Structured UI test report format with coverage and failure tracking
- Sprint workflow skill for phase lifecycle context
- API contract skill for shared interface design
- Sprint plugin README for agent architecture overview
