/flow-next

name: flow-next
description: "Manage .flow/ tasks and specs. Triggers: 'show me my tasks', 'list specs', 'what tasks are there', 'add a task', 'create task', 'what's ready', 'task status', 'show fn-1-add-oauth'. NOT for /flow-next:plan or /flow-next:work."

Flow-Next Task Management

Quick task operations in `.flow/`. For planning features use `/flow-next:plan`, for executing use `/flow-next:work`.

Preamble

**CRITICAL: flowctl is BUNDLED — NOT installed globally.** `which flowctl` will fail (expected). Define once; subsequent blocks use `$FLOWCTL`:

FLOWCTL="${DROID_PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}/scripts/flowctl"
[ -x "$FLOWCTL" ] || FLOWCTL="<plugin-root>/scripts/flowctl"   # <plugin-root> = the directory two levels above this skill's SKILL.md file (the harness gave you that file's absolute path when the skill loaded); substitute it literally
[ -x "$FLOWCTL" ] || FLOWCTL=".flow/bin/flowctl"

**Discover all commands/options:**

$FLOWCTL --help
$FLOWCTL <command> --help   # e.g., $FLOWCTL task --help

Quick Reference

# Check if .flow exists
$FLOWCTL detect --json

# Initialize (if needed)
$FLOWCTL init --json

# List everything (specs + tasks grouped)
$FLOWCTL list --json

# List all specs
$FLOWCTL specs --json

# List all tasks (or filter by spec/status)
$FLOWCTL tasks --json
$FLOWCTL tasks --spec fn-1-add-oauth --json
$FLOWCTL tasks --status todo --json

# View spec with all tasks
$FLOWCTL show fn-1-add-oauth --json
$FLOWCTL cat fn-1-add-oauth              # Spec markdown

# View single task
$FLOWCTL show fn-1-add-oauth.2 --json
$FLOWCTL cat fn-1-add-oauth.2            # Task spec

# What's ready to work on?
$FLOWCTL ready --spec fn-1-add-oauth --json

# Create task under existing spec
$FLOWCTL task create --spec fn-1-add-oauth --title "Fix bug X" --json

# Set task description and acceptance (combined, fewer writes; unique per-task temp paths)
$FLOWCTL task set-spec fn-1-add-oauth.2 --description "${TMPDIR:-/tmp}/flow-desc-fn-1-add-oauth.2.md" --acceptance "${TMPDIR:-/tmp}/flow-accept-fn-1-add-oauth.2.md" --json

# Or use stdin with heredoc (no temp file):
$FLOWCTL task set-description fn-1-add-oauth.2 --file - --json <<'EOF'
Description here
EOF

# Start working on task
$FLOWCTL start fn-1-add-oauth.2 --json

# Mark task done
echo "What was done" > /tmp/summary.md
echo '{"commits":["abc123"],"tests":["npm test"],"prs":[]}' > /tmp/evidence.json
$FLOWCTL done fn-1-add-oauth.2 --summary-file /tmp/summary.md --evidence-json /tmp/evidence.json --json

# Validate structure
$FLOWCTL validate --spec fn-1-add-oauth --json
$FLOWCTL validate --all --json

Common Patterns

"Add a task for X"

1. Find relevant spec:

   # List all specs
   $FLOWCTL specs --json

   # Or show a specific spec to check its scope
   $FLOWCTL show fn-1 --json

2. Create task:

   $FLOWCTL task create --spec fn-N --title "Short title" --json

3. Add description + acceptance (combined):

   # Unique per-task temp paths — written + consumed in this one block
   cat > "${TMPDIR:-/tmp}/flow-desc-fn-N.M.md" << 'EOF'
   **Bug/Feature:** Brief description

   **Details:**
   - Point 1
   - Point 2
   EOF
   cat > "${TMPDIR:-/tmp}/flow-accept-fn-N.M.md" << 'EOF'
   - [ ] Criterion 1
   - [ ] Criterion 2
   EOF
   $FLOWCTL task set-spec fn-N.M --description "${TMPDIR:-/tmp}/flow-desc-fn-N.M.md" --acceptance "${TMPDIR:-/tmp}/flow-accept-fn-N.M.md" --json

"What tasks are there?"

# All specs
$FLOWCTL specs --json

# All tasks
$FLOWCTL tasks --json

# Tasks for specific spec
$FLOWCTL tasks --spec fn-1-add-oauth --json

# Ready tasks for a spec
$FLOWCTL ready --spec fn-1-add-oauth --json

"Show me task X"

$FLOWCTL show fn-1-add-oauth.2 --json   # Metadata
$FLOWCTL cat fn-1-add-oauth.2           # Full spec

(Legacy `fn-1.2` / `fn-1-xxx.2` still works.)

Create new spec (rare - usually via /flow-next:plan)

$FLOWCTL spec create --title "Spec title" --json
# Returns: {"success": true, "id": "fn-N-spec-title", ...}

Close a spec as won't-do

$FLOWCTL spec close fn-1-add-oauth --json

A spec closed **because we decided not to build it** also gets a file in `.flow/memory/declined/<concept-slug>.md`, written directly (agent prose, no flowctl verb): title, the decision in one line, short reasoning, and a `## Prior requests` list opened with today's date and where the request came from. The file already exists → append the dated line under `## Prior requests` and leave the decision as written. The entry body follows the artifact prose contract in [docs/prose.md](../../docs/prose.md); proceed without it when the doc is absent. Without it the concept comes back next quarter with nothing to point at, and the next planner proposes it fresh.

**Only a policy refusal earns a file.** A spec closed as superseded, merged into another spec, already implemented, or obsolete is not a decline — filing it there teaches future planners that shipped or in-flight work is rejected scope. Reopening a declined concept is the user's call alone.

ID Format

• Spec: `fn-N-slug` where slug is derived from title (e.g., `fn-1-add-oauth`, `fn-2-fix-login-bug`)
• Task: `fn-N-slug.M` (e.g., `fn-1-add-oauth.1`, `fn-2-fix-login-bug.2`)

Legacy formats `fn-N` and `fn-N-xxx` (random 3-char suffix) are still supported.

Notes

• Run `$FLOWCTL --help` to discover all commands and options
• **Every write goes through a flowctl subcommand.** A session that edits `.flow/` JSON or task markdown by hand has broken this.
• **Every read comes from `.flow/` state**, via `--json` (`detect`, `list`, `specs`, `tasks`, `show`, `ready`) or `cat` for markdown. An answer assembled from files skimmed by hand has broken this.
• **A task marked complete is closed with `flowctl done` carrying both `--summary-file` and `--evidence-json`.** A bare status flip has broken this.
• **Requests that need real planni