Add Python copy-script installer

.github/workflows/spellcheck.yml

@@ -17,8 +17,8 @@ jobs:
       - name: Check out repository
         uses: actions/checkout@v4
 
-      - name: Install codespell
-        run: python -m pip install --upgrade codespell --break-system-packages
-
       - name: Check Markdown spelling
-        run: codespell $(git ls-files '**/*.md')
+        uses: streetsidesoftware/cspell-action@v6
+        with:
+          config: cspell.json
+          files: "**/*.md"

AGENTS.md

@@ -0,0 +1,24 @@
+# AGENTS.md
+
+## Purpose
+
+This repository packages the `idea-evaluator` skill for multiple agent ecosystems.
+Treat the root files as the canonical project guidance:
+
+- `Intent.md` is the behavior spec for the skill.
+- `README.md` explains the repo layout at a high level.
+- `cspell.json` defines the spellcheck vocabulary used by CI.
+
+## Working Rules
+
+- Read `Intent.md` and `README.md` before making repo-wide changes.
+- Keep the three package folders aligned when a shared description changes.
+- Keep prose concise and consistent across `codex/`, `claude/`, and `openclaw/`.
+- When adding new markdown wording or project terms, update `cspell.json` if CI should accept them.
+- Preserve intentional misspellings only when they are part of a test or a deliberate example.
+- Prefer `apply_patch` for edits and keep changes minimal.
+
+## Validation
+
+- Markdown files are checked in CI with `cspell`.
+- Make sure new or edited `.md` files pass spellcheck before you consider the work done.

Intent.md

@@ -35,6 +35,8 @@ The skill exists to answer:
 
 The evaluation must happen in 3 stages.
 
+Before moving to the next stage, wait until every agent in the current stage has finished its research and returned its output.
+
 ### Stage 1: Broad Scan
 
 Launch 5 subagents.
@@ -110,4 +112,3 @@ The skill is successful if it can:
 - support strong ideas with real reasoning,
 - surface useful market and execution risks,
 - and help the user decide whether the idea deserves an MVP.
-

README.md

@@ -8,6 +8,7 @@ This repository packages the same `idea-evaluator` concept for three different a
 - `codex/idea-evaluator`: Codex-formatted package
 - `claude/idea-evaluator`: Claude Code-formatted package
 - `openclaw/idea-evaluator`: OpenClaw-formatted package
+- `copy-script/`: Python installer and its README
 
 ## What It Does
 
@@ -23,6 +24,11 @@ It is designed to:
 
 - `Intent.md` is the canonical brief for the skill
 - Each package folder contains a `SKILL.md` plus a short README for that ecosystem
+- `copy-script/README.md` explains the installer and its options
+
+## Installation
+
+Run the Python installer in `copy-script/` to copy the package folders into the matching skills directories under your home folder.
 
 ## License
 

claude/idea-evaluator/SKILL.md

@@ -20,6 +20,8 @@ Treat the task like a structured decision engine:
 
 Run the evaluation in 3 stages.
 
+Before moving to the next stage, wait until every agent in the current stage has finished its research and returned its output.
+
 ### Stage 1: Broad Scan
 
 Launch 5 subagents.
@@ -87,4 +89,3 @@ Include:
 - positives
 - negatives
 - MVP question
-

codex/idea-evaluator/SKILL.md

@@ -20,6 +20,8 @@ Treat the task like a structured decision engine:
 
 Run the evaluation in 3 stages.
 
+Before moving to the next stage, wait until every agent in the current stage has finished its research and returned its output.
+
 ### Stage 1: Broad Scan
 
 Launch 5 subagents.
@@ -87,4 +89,3 @@ Include:
 - positives
 - negatives
 - MVP question
-

copy-script/README.md

@@ -0,0 +1,43 @@
+# Copy Script
+
+This folder contains the Python installer for the `idea-evaluator` skill packages.
+
+## What It Does
+
+The script copies each ecosystem package into the matching skills directory in your home folder:
+
+- `codex/idea-evaluator` -> `~/.codex/skills/idea-evaluator`
+- `claude/idea-evaluator` -> `~/.claude/skills/idea-evaluator`
+- `openclaw/idea-evaluator` -> `~/.openclaw/skills/idea-evaluator`
+
+## Interactive Mode
+
+Run the script with no flags in a terminal to pick ecosystems interactively.
+
+It will:
+
+- show the available ecosystems,
+- let you choose one or more,
+- and ask before overwriting existing installs.
+
+## CLI Mode
+
+Use flags when you want a non-interactive run.
+
+Examples:
+
+```bash
+python copy-script/install-skills.py --all
+python copy-script/install-skills.py --ecosystems codex openclaw
+python copy-script/install-skills.py --all --dry-run
+python copy-script/install-skills.py --all --force
+```
+
+## Options
+
+- `--all`: install into every supported ecosystem
+- `--ecosystems`: install into selected ecosystems only
+- `--repo-root`: point the script at a different repository root
+- `--home`: point the script at a different home directory
+- `--dry-run`: print actions without copying files
+- `--force`: skip overwrite confirmation

copy-script/install-skills.py

@@ -0,0 +1,155 @@
+from __future__ import annotations
+
+import argparse
+import shutil
+import sys
+from pathlib import Path
+
+
+ECOSYSTEMS = {
+    "codex": ("codex", ".codex"),
+    "claude": ("claude", ".claude"),
+    "openclaw": ("openclaw", ".openclaw"),
+}
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Copy the idea-evaluator skill packages into local skills directories."
+    )
+    parser.add_argument(
+        "--ecosystems",
+        nargs="+",
+        choices=sorted(ECOSYSTEMS),
+        help="One or more ecosystems to install into. Omit for interactive mode.",
+    )
+    parser.add_argument(
+        "--all",
+        action="store_true",
+        help="Install into all supported ecosystems.",
+    )
+    parser.add_argument(
+        "--repo-root",
+        type=Path,
+        default=Path(__file__).resolve().parent.parent,
+        help="Path to the repository root. Defaults to the parent of copy-script/.",
+    )
+    parser.add_argument(
+        "--home",
+        type=Path,
+        default=Path.home(),
+        help="Home directory that contains the target skills folders.",
+    )
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="Show what would be copied without writing anything.",
+    )
+    parser.add_argument(
+        "--force",
+        action="store_true",
+        help="Overwrite existing installations without prompting.",
+    )
+    return parser
+
+
+def prompt_ecosystems() -> list[str]:
+    options = list(ECOSYSTEMS)
+    print("Select ecosystems to install:")
+    for index, name in enumerate(options, start=1):
+        print(f"  {index}. {name}")
+    print("  a. all")
+
+    choice = input("Enter numbers or names separated by commas [a]: ").strip().lower()
+    if not choice or choice == "a" or choice == "all":
+        return options
+
+    selected: list[str] = []
+    for raw_item in choice.split(","):
+        item = raw_item.strip()
+        if not item:
+            continue
+        if item.isdigit():
+            index = int(item) - 1
+            if index < 0 or index >= len(options):
+                raise SystemExit(f"Invalid ecosystem number: {item}")
+            selected.append(options[index])
+            continue
+        if item not in ECOSYSTEMS:
+            raise SystemExit(f"Unknown ecosystem: {item}")
+        selected.append(item)
+
+    if not selected:
+        raise SystemExit("No ecosystems selected.")
+
+    return list(dict.fromkeys(selected))
+
+
+def prompt_confirmation(message: str) -> bool:
+    choice = input(f"{message} [y/N]: ").strip().lower()
+    return choice in {"y", "yes"}
+
+
+def resolve_targets(args: argparse.Namespace) -> list[str]:
+    if args.all:
+        return list(ECOSYSTEMS)
+
+    if args.ecosystems:
+        return list(dict.fromkeys(args.ecosystems))
+
+    if sys.stdin.isatty():
+        return prompt_ecosystems()
+
+    raise SystemExit("Specify --all or --ecosystems when running non-interactively.")
+
+
+def install_skill(source: Path, destination_root: Path, dry_run: bool) -> Path:
+    destination = destination_root / source.name
+    if dry_run:
+        print(f"Would copy {source} -> {destination}")
+        return destination
+
+    destination_root.mkdir(parents=True, exist_ok=True)
+
+    if destination.exists():
+        if destination.is_dir():
+            shutil.rmtree(destination)
+        else:
+            destination.unlink()
+
+    shutil.copytree(source, destination)
+    print(f"Copied {source} -> {destination}")
+    return destination
+
+
+def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    repo_root = args.repo_root.resolve()
+    home = args.home.expanduser().resolve()
+    selected = resolve_targets(args)
+
+    jobs = []
+    for ecosystem in selected:
+        source_dirname, destination_dirname = ECOSYSTEMS[ecosystem]
+        source = repo_root / source_dirname / "idea-evaluator"
+        destination_root = home / destination_dirname / "skills"
+        jobs.append((ecosystem, source, destination_root))
+
+    if not args.force and not args.dry_run:
+        destinations = [destination_root / "idea-evaluator" for _, _, destination_root in jobs]
+        existing = [path for path in destinations if path.exists()]
+        if existing:
+            print("Existing installations found:")
+            for path in existing:
+                print(f"  {path}")
+            if not sys.stdin.isatty() or not prompt_confirmation("Overwrite them"):
+                raise SystemExit("Cancelled.")
+
+    for _, source, destination_root in jobs:
+        install_skill(source, destination_root, args.dry_run)
+
+
+if __name__ == "__main__":
+    main()

cspell.json

@@ -0,0 +1,23 @@
+{
+  "version": "0.2",
+  "language": "en",
+  "words": [
+    "AGENTS",
+    "Codex",
+    "Claude",
+    "Gitea",
+    "MVP",
+    "OpenClaw",
+    "README",
+    "copy-script",
+    "apply_patch",
+    "cspell",
+    "idea-evaluator",
+    "frontmatter",
+    "subagents",
+    "hyphen-case"
+  ],
+  "ignorePaths": [
+    "LICENSE"
+  ]
+}

openclaw/idea-evaluator/README.md

@@ -6,10 +6,9 @@ This folder contains the OpenClaw version of `idea-evaluator`.
 
 - OpenClaw loads skills from `SKILL.md` files with YAML frontmatter.
 - The skill name lives in frontmatter and should stay lowercase hyphen-case.
-- The body stas concise because it is the operating playbook once the skill triggers.
+- The body stays concise because it is the operating playbook once the skill triggers.
 
 ## Source
 
 - `Intent.md` in the repo root
 - OpenClaw skill loading and frontmatter rules
-

openclaw/idea-evaluator/SKILL.md

@@ -20,6 +20,8 @@ Treat the task like a structured decision engine:
 
 Run the evaluation in 3 stages.
 
+Before moving to the next stage, wait until every agent in the current stage has finished its research and returned its output.
+
 ### Stage 1: Broad Scan
 
 Launch 5 subagents.
@@ -87,4 +89,3 @@ Include:
 - positives
 - negatives
 - MVP question
-