GiustoPiedimonte/agentic-engineering-marketplace
Plugin Claude Code
Agentic engineering tooling: a spec-driven shape/decide/execute/measure/eval workflow that treats work as a graph — fan-out across subagents with adversarial verification — plus README maintenance.
GiustoPiedimonte/agentic-engineering-marketplace
Plugin Claude Code
Spec-driven agentic engineering as graph engineering: shape (pitch) -> decide (ADR) -> execute (delegated cycle with adversarial review) -> measure -> eval, with /graph to fan work out across a fleet of subagents (diamond / router / verifier / cycle). Dark-launch eval-gated. Distilled from a real practice.
GiustoPiedimonte/agentic-engineering-marketplace
Agent
Use this agent to implement ONE approved, closed-scope cycle and open a PR. It is the serialized writer — never run two executors on overlapping files. Context: A pitch has been shaped and approved. user: "Ship the rate-limiter pitch." assistant: "I'll delegate this closed-scope cycle to the executor agent, which will…
GiustoPiedimonte/agentic-engineering-marketplace
Agent
Use this agent to turn a measure-gated decision into a data-backed verdict. Runs read-only checks/queries (DB, logs, metrics) and reports a flip/keep/cut verdict with the data. Never writes code or data. Context: A decision is blocked on "do users actually hit this path?" user: "Before we cut the feature, check if…
GiustoPiedimonte/agentic-engineering-marketplace
Agent
Use this agent for read-only research — codebase investigation, library/API docs, papers, competitor source — feeding shaping/ADR decisions with grounded, cited evidence. Fan-out safe: run several in parallel. Never edits project code. Context: Shaping a pitch that needs grounding. user: "How does our auth handle…
GiustoPiedimonte/agentic-engineering-marketplace
Agent
Use this agent as an adversarial pre-merge gate after an executor opens a PR, before merging. Read-only on code. Returns MERGE / ADJUST / REJECT with specifics. Default to skepticism — find the reason NOT to merge. Context: An executor just opened a PR. user: "Review the PR before I merge." assistant: "I'll run the…
GiustoPiedimonte/agentic-engineering-marketplace
Agent
Use this agent as a general adversarial verifier node in a graph: given ONE finding/claim/answer, try to kill it and return a real/not-real verdict with reasoning. Fan-out safe — run N in parallel (or with distinct lenses) as a gate before a finding is allowed downstream. Read-only. Distinct from reviewer (which…
GiustoPiedimonte/agentic-engineering-marketplace
Hook
Runs after a tool call finishes for Edit, Write and MultiEdit tool calls, executing verify-refs.py with --fix and --no-install (2 commands). From GiustoPiedimonte/agentic-engineering-marketplace.
GiustoPiedimonte/agentic-engineering-marketplace
Hook
Runs when the agent finishes a response, running an inline shell check. From GiustoPiedimonte/agentic-engineering-marketplace.
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to record an architecture decision. Trigger when the user says "record this decision", "write an ADR", "we decided to...", "log why we chose X", or when a consequential, hard-to-reverse technical choice is made. Appends a dated, immutable decision block to the decision log.
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to make eval the unit of progress: systematize how a feature or agent is judged. Trigger with "how do we eval this", "build an eval harness", "why is the agent failing", "error analysis", "are we regressing", or before flipping an eval-gated dark launch. Builds evals from REAL failures (not…
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to design or run a task as a graph instead of a linear chain — fan out independent work across a fleet of subagents, verify findings, and converge. Trigger with "run this as a graph", "fan this out", "parallelize this", "orchestrate this with subagents", "audit every X", "sweep the codebase…
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to unblock a decision that needs data. Trigger with "measure this", "do users actually...", "check the numbers before we cut/keep", "did the change land". Runs a read-only measurement and returns a data-backed flip/keep/cut verdict — never guesses, never writes.
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to shape a feature into a written spec before any code. Trigger when the user says "shape this", "write a pitch", "spec this out", "let's plan ", or starts non-trivial work without an agreed approach. Produces a Shape Up style pitch (the source of truth) via interview.
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
Execute an approved pitch/spec as a delegated, closed-scope cycle with verification and adversarial review. Invoke it yourself with /ship — a cycle spawns a writer and opens a PR, so it starts when you say so. Enforces the execution playbook: right-size gate, pre-spawn filter, doc-bundle, PR format, review.
GiustoPiedimonte/agentic-engineering-marketplace
Plugin Claude Code
Make a public GitHub repo well-made and honest. /readme audits and elevates a README — deriving the hero, voice, badges, and sections from the PROJECT's own identity (fidelity gate, propose-and-confirm), with free-form targeted edits; /opensource adds the community-health files, an honest CI gate, and the right repo…
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to make a repository a well-made public/open-source project. Trigger with "make this repo public-ready", "set up open source", "add a code of conduct / contributing / security policy", "add issue and PR templates", "add CI", "fix the repo settings", or "get this ready to share". Adds the…
GiustoPiedimonte/agentic-engineering-marketplace
Skill Claude CodeCodex
This skill should be used to audit OR elevate the README of a public repository. Trigger with "improve the README", "make the readme high quality", "elevate this readme", "audit the readme", "add badges", "is our README good", or before publishing/releasing a repo. Elevate rebuilds a thin/generic README to a polished…