Workflows

When to use

User wants to report a bug or problem as a GitHub issue with automated reproduction, scoped to the FIT tech stack.

Triggers

Does

Classifies the problem area against the FIT stack (TypeScript, Python, NextJS, n8n, MkDocs, GitHub Actions). Gathers context in parallel — templates, duplicate check, repo metadata. Attempts to reproduce with stack-appropriate tools. Will not create an issue it cannot reproduce. Creates a structured GitHub issue with full evidence and labels.

Not for

Feature requests, enhancements, or non-bug work. Only for bugs that can be reproduced.

When to use

Default FIT issue-fixing workflow. Sonnet 4.6 + medium effort. ~80% of work — well-scoped bugs, small features, anything you'd describe in 1-2 sentences.

Triggers

Does

Ten-phase DAG: classify the issue, web research, investigate or plan (gated by classification), implement, validate, create draft PR using the repo's PR template, run smart review (code review + conditional error-handling/test-coverage/comment-quality/docs-impact agents), self-fix all findings, simplify changed code, report results back to the issue.

Not for

Architectural changes, ambiguous root-cause investigations, governance graph layer work (use fit-fix-issue-pro). Comprehensive multi-agent reviews. Questions about issues. CI failures. PR reviews on their own.

When to use

Heavyweight version of fit-fix-issue. Opus 4.6 (pinned) + medium effort. ~20% of work — architectural changes, ambiguous root-cause, multi-file refactors, or anything touching the governance graph layer (#51 graph traversal, #52 graph audit).

Triggers

Does

Same ten-phase DAG as fit-fix-issue, but pinned to Opus 4.6 for deeper reasoning at the cost of higher token spend. Same review, self-fix, simplify, and report behavior.

Not for

Trivial fixes (use fit-fix-issue). Comprehensive multi-agent reviews (use archon-issue-review-full). Questions about issues. CI failures. PR reviews on their own.

When to use

Run a repo task in the n8n-builder project against the demo n8n environment — testing or validating changes before promoting to dev.

Triggers

Does

Pins the run to the demo env via set-target-env. Validates demo credentials. Verifies n8n MCP connection. Executes the task in a loop (up to 8 iterations) with strict env isolation. Runs lifecycle precheck. Summarizes and persists the run report.

Not for

Dev or prod environment tasks. Cross-environment changes. Tasks that should not touch the live demo instance.

When to use

Run a repo task in the n8n-builder project against the dev n8n environment — the standard development target for building and iterating on workflows.

Triggers

Does

Pins the run to the dev env via set-target-env. Validates dev credentials. Verifies n8n MCP connection. Executes the task in a loop (up to 8 iterations) with strict env isolation. Runs lifecycle precheck. Summarizes and persists the run report.

Not for

Demo or prod environment tasks. Cross-environment changes. Anything that must not modify the dev n8n instance.

When to use

Run a repo task in the n8n-builder project against the production n8n environment. Use sparingly — prefer dev or demo for anything experimental.

Triggers

Does

Pins the run to the prod env. Validates prod credentials. Runs a prod read-only safety check before touching anything. Verifies n8n MCP connection. Executes the task in a loop (up to 8 iterations) with strict env isolation and conservative validation. Runs lifecycle precheck. Summarizes and persists the run report.

Not for

Dev or demo environment tasks. Exploratory or experimental work — always validate on dev first.

When to use

Audit the n8n environment configuration for a FIT repo against required variables, missing secrets, and consistency rules.

Triggers

Does

Runs the deterministic fit-solutions n8n-env-audit Python script against a target repo and env file, then writes JSON and markdown audit artifacts to the run artifacts directory. Read-only — surfaces findings, fixes nothing.

Not for

Mutating configuration. Fixing issues found in the audit. Auditing non-n8n environments.

When to use

Cross-repo status dashboard for all FITAutomate repositories — daily or weekly org health check.

Triggers

Does

Collects branch state, worktrees, open issues, open PRs, and divergence from main across all FIT repos via gh CLI and git. Synthesizes a scannable markdown dashboard with status table, open PRs, issue snapshot, active worktrees, local branch overview. Persists the report to the configured reports directory.

Not for

Fixing issues. Closing PRs. Modifying any repo state. Individual repo deep-dives. Read-only reporting only.

When to use

Draft and publish a new blog post to the FIT Blog — on demand or on a weekly automated schedule.

Triggers

Does

Fifteen-step pipeline. Classifies the input into one of three modes. Decides single vs series post plan. Loads brand voice and service registry docs. Calibrates writing style against the four most recent posts. Extracts topic and angle from source materials when provided. Generates a hero image via OpenAI Images API. Drafts the post. Validates four editorial rules. Validates MkDocs build. Commits the post and hero image locally for human push.

Not for

Editing existing posts. Bulk-importing external content. Automated publishing without human review of the local commit.