deepflow 0.1.69 → 0.1.71

package/README.md

@@ -77,10 +77,8 @@ You write the specs, then walk away. The AI runs the full pipeline — hypothesi
 ```bash
 # You define WHAT (the specs), the AI figures out HOW, overnight
 
-# Requires Agent Teams (experimental feature)
-export CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
-
-deepflow auto                    # process all specs in specs/
+# Inside Claude Code (requires Agent Teams)
+/df:auto                         # process all specs in specs/
 ```
 
 **What the AI does alone:**
@@ -93,7 +91,7 @@ deepflow auto                    # process all specs in specs/
 7. If rejected: generates new hypotheses, retries (up to max-cycles)
 8. On convergence: verifies (L0-L4 gates), creates PR, merges to main
 
-**What you do:** Write specs (via interactive mode or manually) in `specs/`, run `deepflow auto`, read the morning report at `.deepflow/auto-report.md`. No need to run `/df:plan` first — auto mode promotes plain specs to `doing-*` automatically.
+**What you do:** Write specs (via interactive mode or manually) in `specs/`, run `/df:auto` inside Claude Code, read the report at `.deepflow/auto-report.md`. No need to run `/df:plan` first — auto mode promotes plain specs to `doing-*` automatically.
 
 **How to use:**
 ```bash
@@ -103,15 +101,16 @@ $ claude
 > /df:spec auth          # creates specs/auth.md
 > /exit
 
-# In your terminal — enable agent teams and run auto mode
-$ export CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
-$ deepflow auto
+# Inside Claude Code — run auto mode
+> /df:auto
 
 # Next morning — check what happened
 $ cat .deepflow/auto-report.md
 $ git log --oneline
 ```
 
+**Requires:** `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` in your environment (agent teams is an experimental Claude Code feature).
+
 **Safety:** Never pushes to remote. Failed approaches recorded in `.deepflow/experiments/` and never repeated. Specs validated before processing (malformed specs are skipped).
 
 ### The Boundary
@@ -166,7 +165,7 @@ $ git log --oneline
 ## The Flow (Autonomous)
 
 ```
-deepflow auto
+/df:auto
     | Discover specs (auto-promote, topological sort by depends_on)
     | For each doing-* spec:
     |
@@ -225,7 +224,7 @@ Execution happens in an isolated git worktree:
 
 ## LSP Integration
 
-deepflow automatically enables Claude Code's LSP tools during install, giving agents access to `goToDefinition`, `findReferences`, and `workspaceSymbol` for precise code navigation instead of grep-based searching.
+/df:automatically enables Claude Code's LSP tools during install, giving agents access to `goToDefinition`, `findReferences`, and `workspaceSymbol` for precise code navigation instead of grep-based searching.
 
 - **Global install:** sets `ENABLE_LSP_TOOL=1` in `~/.claude/settings.json`
 - **Project install:** sets it in `.claude/settings.local.json`
@@ -235,7 +234,7 @@ Agents prefer LSP tools when available and fall back to Grep/Glob silently. You'
 
 ## Spec Validation
 
-Specs are validated before downstream consumption by `/df:spec`, `/df:plan`, and `deepflow auto`:
+Specs are validated before downstream consumption by `/df:spec`, `/df:plan`, and `/df:auto`:
 
 - **Hard invariants** (block on failure): required sections present, REQ-N prefixes, checkbox ACs, no duplicate IDs
 - **Advisory warnings** (warn interactively, block in auto mode): long specs, orphaned requirements, excessive technical notes
@@ -263,7 +262,7 @@ Statusline shows context usage. At >=50%:
 | `/df:consolidate` | Deduplicate and clean up decisions.md |
 | `/df:resume` | Session continuity briefing |
 | `/df:update` | Update deepflow to latest |
-| `deepflow auto` | Autonomous overnight execution (no human needed) |
+| `/df:auto` | Autonomous execution via agent teams (no human needed) |
 
 ## File Structure
 

package/bin/install.js

@@ -10,25 +10,15 @@ const os = require('os');
 const readline = require('readline');
 const { execFileSync } = require('child_process');
 
-// Subcommand routing: `deepflow auto [...]` -> claude --agent .claude/agents/deepflow-auto.md
+// Legacy subcommand: `deepflow auto` is now `/df:auto` inside Claude Code
 if (process.argv[2] === 'auto') {
-  if (!process.env.CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS) {
-    console.error('Error: CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS environment variable is not set.');
-    console.error('');
-    console.error('The `deepflow auto` command now uses Claude Code Agent Teams.');
-    console.error('To enable it, set the environment variable before running:');
-    console.error('');
-    console.error('  export CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1');
-    console.error('  deepflow auto');
-    console.error('');
-    process.exit(1);
-  }
-  try {
-    execFileSync('claude', ['--agent', '.claude/agents/deepflow-auto.md', '-p', 'Run the full autonomous cycle. Process all doing-* specs.', ...process.argv.slice(3)], { stdio: 'inherit' });
-  } catch (e) {
-    process.exit(e.status || 1);
-  }
-  process.exit(0);
+  console.error('`deepflow auto` has moved inside Claude Code for better visibility.');
+  console.error('');
+  console.error('Usage:');
+  console.error('  1. Open Claude Code: claude');
+  console.error('  2. Run: /df:auto');
+  console.error('');
+  process.exit(1);
 }
 
 // Colors
@@ -196,9 +186,9 @@ async function main() {
   console.log(`${c.green}Installation complete!${c.reset}`);
   console.log('');
   console.log(`Installed to ${c.cyan}${CLAUDE_DIR}${c.reset}:`);
-  console.log('  commands/df/     — /df:discover, /df:debate, /df:spec, /df:plan, /df:execute, /df:verify, /df:note, /df:resume, /df:update');
+  console.log('  commands/df/     — /df:discover, /df:debate, /df:spec, /df:plan, /df:execute, /df:verify, /df:auto, /df:note, /df:resume, /df:update');
   console.log('  skills/          — gap-discovery, atomic-commits, code-completeness');
-  console.log('  agents/          — reasoner, deepflow-auto (autonomous overnight execution)');
+  console.log('  agents/          — reasoner, deepflow-auto (/df:auto — autonomous execution)');
   if (level === 'global') {
     console.log('  hooks/           — statusline, update checker');
   }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "deepflow",
-  "version": "0.1.69",
+  "version": "0.1.71",
   "description": "Stay in flow state - lightweight spec-driven task orchestration for Claude Code",
   "keywords": [
     "claude",

package/src/commands/df/auto.md

@@ -0,0 +1,16 @@
+# /df:auto — Autonomous Mode
+
+Run the full autonomous cycle via agent teams. Auto-promotes unprefixed specs to `doing-*`, then processes all `doing-*` specs through every phase: discover, pre-check, hypothesize, spike, implement, select, verify, PR, report.
+
+## Usage
+```
+/df:auto              # process all specs
+```
+
+## Behavior
+
+Load and execute the lead agent at `.claude/agents/deepflow-auto.md`.
+
+Run the full autonomous cycle now. Auto-promote unprefixed specs to `doing-*`, then process all `doing-*` specs through every phase. Do not ask questions — act autonomously.
+
+Output progress as each phase completes. Generate `.deepflow/auto-report.md` at the end.