From code-forge
Dispatches one agent per independent problem domain for concurrent investigation and fixing. Use when facing 2+ unrelated failures that share no state or dependencies.
How this skill is triggered — by the user, by Claude, or both
Slash command
/code-forge:parallelThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
@../shared/execution-entrypoint.md
@../shared/execution-entrypoint.md
For this skill: start at the first executable step. If you catch yourself about to say "falling back to manual handling", STOP and go to the indicated step.
Dispatch independent sub-agents to solve multiple unrelated problems concurrently.
For task execution within a feature: Use /code-forge:impl which has built-in parallel task support.
One agent per independent problem domain. Let them work concurrently.
List all problems. For each, note:
Build a dependency matrix:
Problem A ←→ Problem B: independent? (yes/no, why)
Problem A ←→ Problem C: independent? (yes/no, why)
Problem B ←→ Problem C: independent? (yes/no, why)
Group dependent problems together. Each independent group gets its own agent.
For each independent problem group, launch a Agent(subagent_type="general-purpose"):
Agent prompt structure:
You are investigating and fixing: {problem description}
## Scope
- Files involved: {list}
- Error messages: {paste}
- Expected behavior: {description}
## Constraints
- Only modify files in your scope
- Use code-forge:tdd methodology (write failing test first)
- Use code-forge:debug methodology (root cause before fix)
## Output Required
- Root cause (1-2 sentences)
- Files changed (list)
- Test results (pass/fail counts)
- Summary of fix (1-2 sentences)
CRITICAL: Launch all agents in a single message using multiple Agent tool calls. This enables true parallel execution.
After all agents complete:
Parallel dispatch complete: {N} agents, {M} problems resolved
Agent 1: {problem} → {root cause} → {status}
Agent 2: {problem} → {root cause} → {status}
Agent 3: {problem} → {root cause} → {status}
Full test suite: {pass}/{total} passing
3 test files failing: auth.test.ts, payment.test.ts, email.test.ts
Independence check:
auth ←→ payment: independent (different modules, no shared state)
auth ←→ email: independent (different modules)
payment ←→ email: independent
Dispatch 3 agents in a single message:
Agent 1: "Fix auth.test.ts — files: src/auth/, error: missing token validation"
Agent 2: "Fix payment.test.ts — files: src/payment/, error: decimal precision"
Agent 3: "Fix email.test.ts — files: src/email/, error: template not found"
Results: 3/3 resolved, full suite 128/128 passing
npx claudepluginhub tercel/tercel-claude-plugins --plugin code-forgeGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.