claudecodex

agentic-workflow

Spec-durable agentic development workflow: artifact loading, per-turn standing-rule re-injection, distillation gating, and the elicitation/audit/spike/retrospective/entropy skills. Implements agentic-dev-workflow.md.

0 on GitHubby Dennis Ev0.2.1updated 4 months agoCoding
/

Plugin agentic-workflow from marketplace agent-workflows in destenson/agent-workflows. Commands read from the repo's manifest — 1 month ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (10)

  • acceptance-stories

    Decompose a written SPEC into discrete user stories, each carrying acceptance criteria and concrete test scenarios, cross-linked back to the spec and the assumptions it depends on. Use after a spec exists and has been assumption-audited, before implementation, to turn "what the system should do" int

  • assumption-audit

    Extract every premise a spec takes for granted, rank by cost-if-false, and write ASSUMPTIONS.md. Run as a fresh instance with the spec as the only input, before implementation. Use after a spec or PRP is drafted.

  • bug-investigation

    Investigate observed wrong behavior, reproduce-first, and record the understanding (ruled-out hypotheses, root cause) — not just the diff. Use when starting any non-trivial bug, as distinct from implementing a spec item.

  • calibrate

    Reason under uncertainty by tracking graded confidence and updating it on evidence. Use when diagnosing a bug, choosing between approaches, estimating risk, or deciding whether to gather more information before acting — anytime a belief is uncertain and tool calls can sharpen it. Trigger on "diagnos

  • design-interview

    Elicit a spec by interviewing the user as a domain expert, then draft SPEC.md / PRPs from their answers. Use at the start of a new project or feature, before any implementation, when requirements are not yet written down.

  • entropy-pass

    Run a deletion-only refactor session — remove code, consolidate duplicates, reduce special cases, with zero new behavior and all validation gates passing. Also prune the journals. Use on a cadence (e.g. every N PRPs) or when entropy has accumulated.

  • fit-review

    Review a working change for approach-fit — was the bug fixed, or the feature built, the right way *for this codebase*? Catches reinvention of existing code, silently broken contracts, symptom-not-cause fixes, unbidden abstraction, and incoherence with local patterns. For bugfixes, verifies the bug i

  • project-docs-init

    Scaffold the four durable workflow artifacts (SPEC, ASSUMPTIONS, DECISIONS, LESSONS), seeding the SPEC framing from the project where it can be discovered. Use when the user asks to initialize, scaffold, or set up the project's durable workflow docs.

  • session-retrospective

    Run the end-of-session distillation by hand — capture dead ends in LESSONS.md and non-obvious choices in DECISIONS.md. Use before wrapping up a session, or when the Stop-hook distillation prompt fires.

  • spike

    Build a throwaway implementation whose only purpose is to validate or falsify one high-risk assumption by making real contact with the data, library, or environment. Use for any high-cost unvalidated assumption before committing to the design.

Hooks (4)

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/scripts/session-start.sh

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/scripts/standing-rules.sh

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/scripts/complexity-gate.sh

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/scripts/distillation-gate.sh

View source on GitHub

More from destenson/agent-workflows

All 9 plugins →

Self-contained base research workflow: loads the durable research artifacts (commonly abstract / proposal / experiments / results) from research/ at session start, re-injects research standing rules per turn, gates session exit on a distillation decision, and ships the research-init / problem-statement / literature-scan / gap-analysis / generate-research-proposal / experiment-design / preregistration / results-memo / negative-results / decision-memo skills.

13 skills · hooks

0

Self-contained workflow for administering and troubleshooting remote systems (Linux/SSH, cloud, containers/k8s, appliances): loads FLEET.md plus the INCIDENTS / CHANGELOG / RUNBOOKS journals at session start, re-injects ops standing rules per turn, flags destructive Bash commands with an out-of-band audit trail, gates session exit on a journaling decision, and ships the fleet-init / fleet-status / troubleshooting-loop / fleet-onboard / incident-record / runbook-capture skills.

9 skills · hooks

0

c-suite

claudecodex

Convenes executive-officer review lenses (CEO / CFO / COO / CTO / CMO / CLO) on a single cross-domain decision and preserves their disagreement instead of averaging it into false consensus. Keeps a small governance spine (CHARTER / STRATEGY / BOARD) as the project's executive memory, ships the c-suite-init and board-review convening workflows plus per-officer review skills, and gates session exit on recording any board decision (and its dissent) into BOARD.md.

8 skills · hooks

0

embedded-target

claudecodex

Extension of agentic-workflow for remote embedded Linux targets (e.g. Jetson): loads DEVICE.md at session start, scaffolds the build-deploy-run-collect harness, repro library, fault-injection and headers-only capture templates, and ships the reproducible / instrument-and-wait debug-loop skills. Implements embedded-target-workflow.md.

6 skills · hooks

0

Hardens a working prototype into a pre-release product. Keeps a HARDENING.md gap ledger (release bar plus the itemized open gaps), loaded at session start together with the hardening standing rules, plus a non-blocking reminder to keep the ledger current. Ships assess-prototype (build the ledger), define-release-target (set the bar and prioritize, extending SPEC.md if present), and convert-prototype (work the ledger down one item per commit). Activates only once a HARDENING.md exists.

3 skills · hooks

0

handoff

claudecodex

Standalone session-handoff plugin: loads a pending HANDOFF.md at session start so the next session resumes where the last left off, gives a non-blocking reminder to record unfinished and newly-discovered work before wrapping up, and ships the handoff skill that writes, updates, and clears the rolling handoff. Enable or disable independently of the other workflows.

2 skills · hooks

0

Similar plugins

Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).

39 skills

mattpocock/skills 323,108

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

claude-mem

claudecodex

Persistent memory for Claude Code: the next session starts where the last one left off.

29 skills · 1 MCP server · hooks

thedotmack/claude-mem 116,618

career-ops

claudecodex

Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications

5 skills · hooks

career-ops-hq/career-ops 82,253

mem0

claudecodex

Cross-session memory and token savings for coding agents.

6 skills · 1 agent · 1 MCP server · hooks

mem0ai/mem0 68,893

context7

claudecodex

Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.

1 MCP server

upstash/context7 64,709

Maintain agentic-workflow?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins9 plugins