feat: bash setup script, planner agent with disallowedTools, simplified skills
This commit is contained in:
112
agents/planner.md
Normal file
112
agents/planner.md
Normal file
@@ -0,0 +1,112 @@
|
||||
---
|
||||
name: planner
|
||||
description: Generates prd.json and sprint contracts from a spec. Can only write to .loop/ directory.
|
||||
model: sonnet
|
||||
maxTurns: 30
|
||||
disallowedTools:
|
||||
- Bash
|
||||
---
|
||||
|
||||
You are a planner agent for the agent loop harness. Your job is to decompose a feature spec into user stories and sprint contracts.
|
||||
|
||||
## CONSTRAINTS
|
||||
|
||||
- You may ONLY write files inside the `.loop/` directory
|
||||
- You may NOT write any project source code (.js, .ts, .py, .go, .rs, .html, .css, etc.)
|
||||
- You may NOT run bash commands
|
||||
- You may NOT start implementing features
|
||||
- You produce prd.json and contracts, then STOP
|
||||
|
||||
## YOUR TASK
|
||||
|
||||
You will be given a feature spec or description. Decompose it into stories.
|
||||
|
||||
### prd.json Schema (EXACT — do not deviate)
|
||||
|
||||
Write `.loop/prd.json`:
|
||||
|
||||
```json
|
||||
{
|
||||
"project": "<project name>",
|
||||
"branchName": "loop/<feature-slug>",
|
||||
"description": "<one-line description>",
|
||||
"userStories": [
|
||||
{
|
||||
"id": "US-001",
|
||||
"title": "Short title",
|
||||
"description": "What this delivers",
|
||||
"acceptanceCriteria": [
|
||||
"Specific verifiable criterion"
|
||||
],
|
||||
"priority": 1,
|
||||
"passes": false,
|
||||
"notes": "",
|
||||
"rejections": 0
|
||||
}
|
||||
]
|
||||
}
|
||||
```
|
||||
|
||||
Rules:
|
||||
- Top-level array MUST be `userStories`. NOT `sprints`, `stories`, `tasks`.
|
||||
- IDs: US-001, US-002, etc.
|
||||
- `passes`: false. `notes`: "". `rejections`: 0.
|
||||
- Each priority unique (1 = highest).
|
||||
- Each story fits one agent context window (1-3 files changed).
|
||||
- Acceptance criteria: independently verifiable. Not "works well" — "returns X when Y".
|
||||
|
||||
### Sprint Contracts
|
||||
|
||||
Create `.loop/contracts/` directory. For each story, write `.loop/contracts/{id}.contract.md`:
|
||||
|
||||
```markdown
|
||||
# Sprint Contract: {id} — {title}
|
||||
|
||||
## What Will Be Built
|
||||
Concrete deliverable.
|
||||
|
||||
## Done Conditions
|
||||
- [ ] Specific testable condition
|
||||
- [ ] All acceptance criteria met
|
||||
|
||||
## Evaluation Criteria
|
||||
- [ ] What evaluator checks
|
||||
- [ ] No regressions
|
||||
|
||||
## Out of Scope
|
||||
- Not in this story
|
||||
|
||||
## Key Files
|
||||
- path/to/file — what changes
|
||||
|
||||
## Dependencies
|
||||
- Depends on: [IDs or "none"]
|
||||
- Blocks: [IDs or "none"]
|
||||
```
|
||||
|
||||
### Progress File
|
||||
|
||||
Write `.loop/progress.md`:
|
||||
|
||||
```markdown
|
||||
# Progress
|
||||
|
||||
## Codebase Patterns
|
||||
|
||||
- [Patterns from analysis]
|
||||
|
||||
---
|
||||
|
||||
## Session Log
|
||||
|
||||
### Planning Session
|
||||
Date: {today}
|
||||
|
||||
**PRD created:** {N} stories for "{description}"
|
||||
|
||||
---
|
||||
```
|
||||
|
||||
## WHEN DONE
|
||||
|
||||
List all stories with their IDs and titles. Then STOP. Do not implement anything.
|
||||
124
setup.sh
Executable file
124
setup.sh
Executable file
@@ -0,0 +1,124 @@
|
||||
#!/bin/bash
|
||||
# Agent Loop — project setup script
|
||||
# Scaffolds .loop/ directory and generates config.json.
|
||||
# Called by /agent-loop:init or run directly.
|
||||
#
|
||||
# Usage:
|
||||
# setup.sh <mode> # mode: implement, explore, or fix
|
||||
# setup.sh implement # scaffold + config for implement mode
|
||||
# setup.sh # defaults to implement
|
||||
|
||||
set -euo pipefail
|
||||
|
||||
MODE="${1:-implement}"
|
||||
|
||||
# --- Validate mode ---
|
||||
if [[ ! "$MODE" =~ ^(implement|explore|fix)$ ]]; then
|
||||
echo "[setup] ERROR: Invalid mode '$MODE'. Must be: implement, explore, fix"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
# --- Find harness source ---
|
||||
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
|
||||
HARNESS_SRC=""
|
||||
|
||||
# Option 1: Running from the plugin cache or repo (script is next to prompts/)
|
||||
if [ -d "$SCRIPT_DIR/prompts" ]; then
|
||||
HARNESS_SRC="$SCRIPT_DIR"
|
||||
fi
|
||||
|
||||
# Option 2: Plugin cache (glob for any version)
|
||||
if [ -z "$HARNESS_SRC" ]; then
|
||||
HARNESS_SRC=$(ls -d ~/.claude/plugins/cache/agent-loop/agent-loop/*/prompts/.. 2>/dev/null | head -1)
|
||||
fi
|
||||
|
||||
# Option 3: Global install
|
||||
if [ -z "$HARNESS_SRC" ] && [ -d "$HOME/.claude/loop/prompts" ]; then
|
||||
HARNESS_SRC="$HOME/.claude/loop"
|
||||
fi
|
||||
|
||||
if [ -z "$HARNESS_SRC" ]; then
|
||||
echo "[setup] ERROR: Cannot find agent-loop harness files."
|
||||
echo "[setup] Install the plugin: /plugin install agent-loop@agent-loop"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
echo "[setup] Harness source: $HARNESS_SRC"
|
||||
|
||||
# --- Scaffold .loop/ ---
|
||||
PROJECT_ROOT="$(pwd)"
|
||||
LOOP_DIR="$PROJECT_ROOT/.loop"
|
||||
|
||||
if [ -d "$LOOP_DIR" ] && [ -f "$LOOP_DIR/prd.json" ]; then
|
||||
echo "[setup] .loop/ already exists with prd.json."
|
||||
echo "[setup] To re-initialize, delete .loop/ first: rm -rf .loop"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
mkdir -p "$LOOP_DIR"
|
||||
|
||||
# Copy harness files
|
||||
cp -r "$HARNESS_SRC/prompts" "$LOOP_DIR/"
|
||||
cp -r "$HARNESS_SRC/templates" "$LOOP_DIR/"
|
||||
cp -r "$HARNESS_SRC/lib" "$LOOP_DIR/"
|
||||
cp "$HARNESS_SRC/loop.sh" "$LOOP_DIR/"
|
||||
chmod +x "$LOOP_DIR/loop.sh"
|
||||
|
||||
# Verify critical files
|
||||
for f in prompts/generator/_base.md prompts/evaluator/_base.md templates/progress.md.template lib/state.sh loop.sh; do
|
||||
if [ ! -f "$LOOP_DIR/$f" ]; then
|
||||
echo "[setup] ERROR: Missing $LOOP_DIR/$f — harness copy failed"
|
||||
exit 1
|
||||
fi
|
||||
done
|
||||
|
||||
# --- Create .gitignore ---
|
||||
cat > "$LOOP_DIR/.gitignore" << 'GITIGNORE'
|
||||
prd.json
|
||||
progress.md
|
||||
progress-archive.md
|
||||
config.json
|
||||
init.sh
|
||||
contracts/
|
||||
triage/
|
||||
archive/
|
||||
.archive-staging/
|
||||
.last-branch
|
||||
.loop.lock
|
||||
GITIGNORE
|
||||
|
||||
# --- Generate config.json ---
|
||||
cat > "$LOOP_DIR/config.json" << EOF
|
||||
{
|
||||
"tool": "claude",
|
||||
"mode": "$MODE",
|
||||
"maxIterations": 20,
|
||||
"skipEval": false,
|
||||
"evalRetries": 2,
|
||||
"autoHooks": true,
|
||||
"branchPrefix": "loop/",
|
||||
"scopeBudgets": {
|
||||
"explore": { "maxFilesToRead": 15, "maxLinesToWrite": 0, "maxFilesToModify": 0 },
|
||||
"implement": { "maxFilesToRead": 50, "maxLinesToWrite": 500, "maxFilesToModify": 10 },
|
||||
"fix": { "maxFilesToRead": 30, "maxLinesToWrite": 200, "maxFilesToModify": 5 }
|
||||
}
|
||||
}
|
||||
EOF
|
||||
|
||||
# --- Generate init.sh stub ---
|
||||
cat > "$LOOP_DIR/init.sh" << 'INITSH'
|
||||
#!/bin/bash
|
||||
set -euo pipefail
|
||||
echo "[init] Environment ready."
|
||||
INITSH
|
||||
chmod +x "$LOOP_DIR/init.sh"
|
||||
|
||||
# --- Done ---
|
||||
echo ""
|
||||
echo "[setup] .loop/ scaffolded successfully."
|
||||
echo "[setup] Mode: $MODE"
|
||||
echo "[setup] Config: .loop/config.json"
|
||||
echo ""
|
||||
echo "Next steps (in Claude Code):"
|
||||
echo " /agent-loop:plan # Generate stories from your spec or description"
|
||||
echo ""
|
||||
@@ -1,102 +1,35 @@
|
||||
---
|
||||
name: init
|
||||
description: "Infrastructure setup: scaffold .loop/ directory with harness files and generate config.json. Does NOT plan or write code."
|
||||
description: "Run the setup script to scaffold .loop/ directory. Does not plan features or write code."
|
||||
---
|
||||
|
||||
# /init — Scaffold the Agent Loop
|
||||
|
||||
Create the `.loop/` directory and generate `config.json`. That's it. This skill does NOT plan features, generate stories, or write any project source code.
|
||||
Run the setup script to create `.loop/` with harness files and config. This skill does ONE thing: run a bash command.
|
||||
|
||||
## STOP: What This Skill Does NOT Do
|
||||
## Instructions
|
||||
|
||||
- Does NOT read or use any spec files
|
||||
- Does NOT generate prd.json or stories
|
||||
- Does NOT write any source code
|
||||
- Does NOT invoke /agent-loop:plan
|
||||
- Does NOT start the loop
|
||||
1. Ask the user: **Mode?** (a) Implement (b) Explore (c) Fix — default is Implement.
|
||||
|
||||
## Steps
|
||||
|
||||
### 1. Check if .loop/ already exists
|
||||
|
||||
If yes, ask: "`.loop/` already exists. Re-initialize? (This resets config but keeps prd.json and progress.md)"
|
||||
|
||||
If they say no, stop.
|
||||
|
||||
### 2. Copy harness files
|
||||
|
||||
Run these commands:
|
||||
2. Run the setup script:
|
||||
|
||||
```bash
|
||||
mkdir -p .loop
|
||||
|
||||
# Find plugin harness source
|
||||
HARNESS_SRC=$(ls -d ~/.claude/plugins/cache/agent-loop/agent-loop/*/prompts/.. 2>/dev/null | head -1)
|
||||
|
||||
if [ -n "$HARNESS_SRC" ]; then
|
||||
cp -r "$HARNESS_SRC/prompts" .loop/
|
||||
cp -r "$HARNESS_SRC/templates" .loop/
|
||||
cp -r "$HARNESS_SRC/lib" .loop/
|
||||
cp "$HARNESS_SRC/loop.sh" .loop/
|
||||
chmod +x .loop/loop.sh
|
||||
else
|
||||
echo "ERROR: Could not find agent-loop plugin files"
|
||||
fi
|
||||
bash ~/.claude/plugins/cache/agent-loop/agent-loop/*/setup.sh <mode>
|
||||
```
|
||||
|
||||
Verify `.loop/prompts/generator/_base.md` exists. If not, report the error and stop.
|
||||
If that path doesn't work, try:
|
||||
|
||||
### 3. Create .loop/.gitignore
|
||||
|
||||
```
|
||||
prd.json
|
||||
progress.md
|
||||
progress-archive.md
|
||||
config.json
|
||||
init.sh
|
||||
contracts/
|
||||
triage/
|
||||
archive/
|
||||
.archive-staging/
|
||||
.last-branch
|
||||
.loop.lock
|
||||
```bash
|
||||
bash "$(ls -d ~/.claude/plugins/cache/agent-loop/agent-loop/*/setup.sh 2>/dev/null | head -1)" <mode>
|
||||
```
|
||||
|
||||
### 4. Detect project and generate config
|
||||
3. Show the script output to the user.
|
||||
|
||||
Read the project root to detect tech stack (package.json, go.mod, Cargo.toml, etc.).
|
||||
4. Tell the user:
|
||||
|
||||
Ask the user:
|
||||
> **Mode?** (a) Explore (b) Implement (c) Fix
|
||||
|
||||
Write `.loop/config.json`:
|
||||
|
||||
```json
|
||||
{
|
||||
"tool": "claude",
|
||||
"mode": "<implement|explore|fix>",
|
||||
"maxIterations": 20,
|
||||
"skipEval": false,
|
||||
"evalRetries": 2,
|
||||
"autoHooks": true,
|
||||
"branchPrefix": "loop/",
|
||||
"scopeBudgets": {
|
||||
"explore": { "maxFilesToRead": 15, "maxLinesToWrite": 0, "maxFilesToModify": 0 },
|
||||
"implement": { "maxFilesToRead": 50, "maxLinesToWrite": 500, "maxFilesToModify": 10 },
|
||||
"fix": { "maxFilesToRead": 30, "maxLinesToWrite": 200, "maxFilesToModify": 5 }
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 5. Generate init.sh
|
||||
|
||||
Write `.loop/init.sh` with project-specific setup (dependency install, dev server, test runner). Make it executable.
|
||||
|
||||
### 6. Done
|
||||
|
||||
Tell the user:
|
||||
|
||||
> `.loop/` is ready. Next steps:
|
||||
> `.loop/` is ready. Next:
|
||||
> ```
|
||||
> /agent-loop:plan # Generate stories from your spec or feature description
|
||||
> /agent-loop:plan
|
||||
> ```
|
||||
|
||||
**That's it. Do not do anything else. Do not read specs. Do not plan. Do not write code.**
|
||||
|
||||
@@ -1,154 +1,63 @@
|
||||
---
|
||||
name: plan
|
||||
description: "Generate prd.json and sprint contracts from a spec or feature description. Requires .loop/ directory. Does NOT write source code or start the loop."
|
||||
description: "Generate prd.json and sprint contracts by dispatching the planner agent. Does not write source code."
|
||||
---
|
||||
|
||||
# /plan — Generate PRD and Sprint Contracts
|
||||
|
||||
Read a spec or feature description, decompose it into stories, and write `.loop/prd.json` and `.loop/contracts/`. This skill does NOT write any project source code or start the loop.
|
||||
Dispatch the planner agent to decompose a spec into stories. The planner agent cannot write source code or run bash commands — it can only write to `.loop/`.
|
||||
|
||||
## STOP: What This Skill Does NOT Do
|
||||
## Instructions
|
||||
|
||||
- Does NOT write any project source code (no .js, .py, .ts, .html, .css, etc.)
|
||||
- Does NOT start the loop or invoke /agent-loop:run
|
||||
- Does NOT make git commits to project code
|
||||
### 1. Check prerequisites
|
||||
|
||||
## Prerequisites
|
||||
Verify `.loop/config.json` exists. If not, tell the user to run `/agent-loop:init` first and stop.
|
||||
|
||||
`.loop/config.json` must exist. If not, tell the user to run `/agent-loop:init` first and stop.
|
||||
### 2. Find the spec
|
||||
|
||||
## Steps
|
||||
|
||||
### 1. Find the feature spec or description
|
||||
|
||||
Check these locations for an existing spec:
|
||||
Check these locations:
|
||||
- `docs/superpowers/specs/*.md`
|
||||
- `docs/specs/*.md`
|
||||
- `SPEC.md`, `PRD.md`, `DESIGN.md` at project root
|
||||
|
||||
If found: "I found a spec at `{path}`. Using it."
|
||||
If found, read the spec content.
|
||||
|
||||
If not found, check if the user passed a description as an argument. If nothing: ask "What do you want to build? 1-3 sentences."
|
||||
If not found, ask: "What do you want to build? 1-3 sentences." Use their answer as the spec.
|
||||
|
||||
### 2. Analyze the project
|
||||
### 3. Read project context
|
||||
|
||||
- Read relevant source directories
|
||||
- Check for existing tests and conventions
|
||||
- Run `git log --oneline -20` if git history exists
|
||||
Read the project root listing and any key config files (package.json, etc.) to understand the tech stack. Read `.loop/config.json` for the mode.
|
||||
|
||||
### 3. Ask 2-3 clarifying questions
|
||||
### 4. Dispatch planner agent
|
||||
|
||||
Only questions where human judgment is needed. Do NOT ask what you can answer from the code or spec.
|
||||
Use the Agent tool to launch the `agent-loop:planner` agent:
|
||||
|
||||
### 4. Generate prd.json
|
||||
|
||||
**CRITICAL: Use this EXACT schema. The loop orchestrator will break if you deviate.**
|
||||
|
||||
Write `.loop/prd.json`:
|
||||
|
||||
```json
|
||||
{
|
||||
"project": "<project name>",
|
||||
"branchName": "loop/<feature-slug>",
|
||||
"description": "<one-line description>",
|
||||
"userStories": [
|
||||
{
|
||||
"id": "US-001",
|
||||
"title": "Short descriptive title",
|
||||
"description": "What this story delivers",
|
||||
"acceptanceCriteria": [
|
||||
"Specific, verifiable criterion",
|
||||
"Another criterion"
|
||||
],
|
||||
"priority": 1,
|
||||
"passes": false,
|
||||
"notes": "",
|
||||
"rejections": 0
|
||||
}
|
||||
]
|
||||
}
|
||||
```
|
||||
Agent(
|
||||
subagent_type: "agent-loop:planner",
|
||||
prompt: "Generate prd.json and sprint contracts for this project.\n\nMode: {mode}\nProject root: {path}\n\nSpec:\n{spec content}\n\nProject tech stack: {detected stack}",
|
||||
description: "Planning: generate stories"
|
||||
)
|
||||
```
|
||||
|
||||
**Rules:**
|
||||
- Top-level array key MUST be `userStories`. Not `sprints`, `stories`, or `tasks`.
|
||||
- IDs: `US-001`, `US-002`, etc.
|
||||
- `passes`: always `false`. `notes`: always `""`. `rejections`: always `0`.
|
||||
- Each priority is unique (1 = highest).
|
||||
- Each story fits in one agent context window (1-3 files changed).
|
||||
- Acceptance criteria must be independently verifiable — not "works well" but "returns X when given Y".
|
||||
### 5. Present results
|
||||
|
||||
### 5. Generate sprint contracts
|
||||
|
||||
Create `.loop/contracts/` directory. For each story, write `.loop/contracts/{id}.contract.md`:
|
||||
|
||||
```markdown
|
||||
# Sprint Contract: {id} — {title}
|
||||
|
||||
## What Will Be Built
|
||||
Concrete deliverable description.
|
||||
|
||||
## Done Conditions
|
||||
- [ ] Specific testable condition
|
||||
- [ ] All acceptance criteria from prd.json met
|
||||
|
||||
## Evaluation Criteria
|
||||
- [ ] What the evaluator checks
|
||||
- [ ] No regressions
|
||||
|
||||
## Out of Scope
|
||||
- Things NOT in this story
|
||||
|
||||
## Key Files
|
||||
- path/to/file — what changes
|
||||
|
||||
## Dependencies
|
||||
- Depends on: [IDs or "none"]
|
||||
- Blocks: [IDs or "none"]
|
||||
```
|
||||
|
||||
### 6. Initialize progress.md
|
||||
|
||||
Write `.loop/progress.md`:
|
||||
|
||||
```markdown
|
||||
# Progress
|
||||
|
||||
## Codebase Patterns
|
||||
|
||||
- [Patterns from analysis]
|
||||
|
||||
---
|
||||
|
||||
## Session Log
|
||||
|
||||
### Planning Session
|
||||
Date: {today}
|
||||
|
||||
**PRD created:** {N} stories for "{description}"
|
||||
|
||||
---
|
||||
```
|
||||
|
||||
### 7. Present for review
|
||||
After the planner finishes, read `.loop/prd.json` and show the user:
|
||||
|
||||
> **Plan Ready — Review Before Running**
|
||||
>
|
||||
> | Stories | Mode | Branch |
|
||||
> |---------|------|--------|
|
||||
> | {N} | {mode} | {branchName} |
|
||||
>
|
||||
> **Stories:**
|
||||
> 1. US-001: {title}
|
||||
> 2. US-002: {title}
|
||||
> ...
|
||||
>
|
||||
> **Review before running:**
|
||||
> **Review:**
|
||||
> - `.loop/prd.json` — stories and acceptance criteria
|
||||
> - `.loop/contracts/` — done conditions per story
|
||||
>
|
||||
> When ready:
|
||||
> Adjust anything, then:
|
||||
> ```
|
||||
> /agent-loop:run
|
||||
> ```
|
||||
|
||||
**STOP HERE. Do NOT start the loop. Wait for the user to run /agent-loop:run explicitly.**
|
||||
**STOP. Do not start the loop. Do not write source code.**
|
||||
|
||||
Reference in New Issue
Block a user