Skip to content

When to Use

Use /flow-next:capture if the requirement is in the current conversation, then plan the generated spec. Skip chart - intent and boundaries are already stateable (signal absent).

Terminal window
/flow-next:capture
/flow-next:plan fn-1
/flow-next:work fn-1

Use /flow-next:capture. It synthesizes the current thread into a source-tagged spec and shows a read-back before writing.

One large idea, still too unclear to capture

Section titled “One large idea, still too unclear to capture”

Use /flow-next:chart. It resolves one decision at a time until the effort is briefable, then hands a briefing package to capture. Chart is optional discovery - never a mandatory stage.

Terminal window
/flow-next:chart multi-tenant billing with unknown pricing and migration risks
# ... resolve decisions until briefable ...
/flow-next:capture

Use /flow-next:guide. It recommends the smallest sufficient route from your starting state and names whether a skip is signal-absent or despite unresolved risk. To see the reasoning worked through for five common shapes — epic, feature, small task, bug, docs chore — read Pipeline Variations.

Use /flow-next:interview. For teams, split the interview:

Terminal window
/flow-next:interview fn-1 --scope=business
/flow-next:interview fn-1 --scope=technical --strategy --docs

Use /flow-next:prospect. It ranks candidate ideas grounded in the repo, strategy, memory, and open specs. After you pick a candidate, chart only if it remains oversized and unclear; otherwise capture.

Use a business interview before the technical pass:

Terminal window
/flow-next:interview fn-1 --scope=business
/flow-next:interview fn-1 --scope=technical

This is the best default for teams where a PO, PM, designer, or support lead owns part of the requirement.

Use the full review stack:

Terminal window
/flow-next:plan-review fn-1
/flow-next:work fn-1
/flow-next:impl-review fn-1
/flow-next:spec-completion-review fn-1
/flow-next:make-pr fn-1

The plan review checks whether the work is safe to start. Implementation review checks the diff. Completion review checks the whole spec after all tasks are done.

Three loops cover the autonomy spectrum — same pipeline, same gates (Going Autonomous has the full picture):

  • Pilot — the build loop. One tick advances one ready spec by one stage; your host’s /loop or /goal drives it from blessed spec to draft PR.
  • Land — the ship loop. Babysits the PRs the build loop opened: CI green, reviews converged, merged, released.
  • Ralph — the hardened harness for fully planned specs: fresh session per iteration, hook-enforced guardrails. Use it only after specs are crisp enough to verify.
Terminal window
/loop 10m /flow-next:pilot # build loop
/loop 30m /flow-next:land # ship loop
/flow-next:ralph-init # hardened harness
scripts/ralph/ralph_once.sh
scripts/ralph/ralph.sh
  • One-line edits where no durable context matters.
  • Throwaway prototypes that will not be reviewed or maintained.
  • Product decisions the agent should not make.
  • Work where nobody will read the spec or receipts.

When in doubt, start with a small spec. Flow-Next adds the most value when ambiguity, coordination, or review cost would otherwise dominate the work.