Workflows

@coleam00

Guided Plan-Implement-Validate development with human-in-the-loop checkpoints. Plan your feature, implement with AI, then validate before committing.

@coleam00

Automatically fix, resolve, or implement a solution for a GitHub issue. Syncs the issue, plans the fix, implements it, and opens a PR.

@coleam00

Full code review of a pull request with automatic fixes. Runs 5 specialized review agents in parallel, synthesizes findings, and auto-fixes critical issues.

@coleam00

Ralph implementation loop — generate or load a PRD, break it into stories, then run Ralph iteratively until all stories are complete.

@coleam00

Turn a freeform prompt (URL, GitHub repo, release notes, topic) into a voiced + animated Remotion video. Three approval gates let you steer the spec, script, and live preview before render. Requires an ElevenLabs API key.

@lamachine

Interactive 8-node workflow that turns a raw idea into BKM-format Work Orders through four AI phases with approval gates between each: understand the idea, scope and approach, risk and decomposition, generate WOs. Output is a directory of self-contained WO files ready to hand to archon-piv-loop.

@lraphael

GitLab counterpart to archon-smart-pr-review. Adaptive code review of a GitLab MR — Haiku classifies which review agents are relevant, runs them in parallel, posts resolvable Discussion threads, and auto-approves on 0 critical findings.

@lraphael

GitLab counterpart to archon-comprehensive-pr-review. Full code review of a GitLab MR — all 5 review agents (code-review, error-handling, test-coverage, comment-quality, docs-impact) run in parallel, posts resolvable Discussion threads, auto-approves on 0 critical findings.

@coleam00

Runs the PIV loop (Plan-Implement-Validate) on a feature or bug behind four human-in-the-loop gates, then evolves the codebase's own AI Layer from what the run learned. Eight phases are adapted from the agentic-coding-course AI Layer. Ends in a draft PR.

@seanrobertwright

Audit a repository for AI-agent readiness. A deterministic script runs 10 checks (agent instructions, README, build/test commands, CI, tests, lint/typecheck, .archon workflows, security scanning, sensitive paths) and computes a score out of 100; Claude then narrates the pre-computed results into a markdown report with a rating and the top 3 fixes ranked by score gain.

@seanrobertwright

Matt Pocock's 'Skills for Real Engineers' (v1.1.0) as six workflows that mount the real SKILL.md files into nodes: spec-to-ship (tracer-bullet tickets -> TDD frontier loop -> two-axis Standards|Spec review -> PR), conservative AFK triage, a diagnosing-bugs pipeline that refuses to theorize without a red feedback loop, an AFK wayfinder frontier, architecture health scans, and repo init. Interactive grilling stays human — the spec issue is the handoff artifact into the AFK back half.

@TheSmokeDev

Point-and-shoot programmatic SEO/GEO page factory. Scan any website, then expand it into hundreds of validated 2,800+ word answer-first pages with hard uniqueness gates and packet-only facts. Never deploys - generation ends at a validation report.

@TheSmokeDev

One visual brief in, a grounded image prompt pack out (optional Codex-rendered bitmaps). Template selection cites a pinned, checksum-verified 511-case style corpus; a deterministic validator fails the run on any citation that does not resolve. Renders are opt-in and every concept ships baked and overlay variants.