E Stpierre Agentic Forge
e-stpierre-agentic-forge is a Claude Code plugin for the Build phase that runs YAML-defined multistep agent workflows with parallel orchestration, conditionals, and retry-based error recovery.
by e-stpierre · github.com/e-stpierre/agentic-forge
Define repeatable agent workflows in YAML with parallel steps, retries, and conditionals instead of one-off prompt chains.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install e-stpierre-agentic-forge@e-stpierre/agentic-forgeBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:e-stpierre/agentic-forge") and the paste-ready config with skillselion.get_install_config("plugin:e-stpierre/agentic-forge").
What it does
Agentic Forge is a Claude Code plugin that packages a YAML-based agentic workflow engine for builders who need structured, repeatable automation instead of fragile ad-hoc prompts. You declare steps, parallelism, conditionals, and retry behavior in configuration, then let the engine coordinate multistep and long-running work with error recovery. It targets indie developers and small teams shipping agent-heavy products—CLIs, internal tools, and API-backed services—who already use Claude Code and want SDLC-style tasks (build, verify, deploy prep) expressed as workflows. Use it when a single skill or MCP call is too narrow and you need orchestration with clear failure handling. It matters because it turns agent work into inspectable, versionable workflow definitions you can iterate in Git rather than re-explaining execution order every session. Complexity is advanced: expect to design graphs, handle partial failures, and tune retries. The repo is early (1 GitHub star at catalog ingest); treat capabilities as beta and validate against your own jobs before production reliance.
Highlights
- YAML-defined multistep workflows with conditional branches and explicit execution logic
- Parallel orchestration for independent steps with error recovery and retry mechanisms
- Supports short tasks and long-running operations in one engine (beta)
- SDLC-oriented task modeling for complex agentic runs beyond single-shot tools
- Single-plugin bundle (pluginCount: 1) from e-stpierre/agentic-forge
Why builders use it
One-off agent prompts break when steps multiply, failures need retries, or work must run in parallel across a real SDLC.
After you register the plugin, you can encode agent tasks as reusable YAML workflows with orchestration and recovery instead of rebuilding execution logic each time.
At a glance
- Type - Plugin in AI Agents.
- Adoption - 0 installs, 1 stars, 0 votes.
FAQ
Who is e-stpierre-agentic-forge for?
It is for Claude Code users who want declarative, multistep agent workflows with parallel execution and retries, especially around development and delivery tasks.
When should I use e-stpierre-agentic-forge?
Use it when a task spans several dependent or parallel agent steps, needs conditional logic, or must survive transient failures with recovery rather than a single chat turn.
How do I add e-stpierre-agentic-forge to my agent?
Install the plugin from the e-stpierre/agentic-forge repository into Claude Code’s plugin flow, then author and run workflows per the plugin’s YAML workflow conventions.
Comments
Share how you use e-stpierre-agentic-forge, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.