script-generator — Claude Code skill: what it does, how to install
script-generator generates optimized, syntax-checked Python, Bash, and PowerShell scripts. It builds and validates the files but never executes them.
MIT 176 installs ★ 632 updated 2026-07-29
npx skills add transilienceai/communitytools --skill script-generator -a claude-codeWhat script-generator does
The skill targets scripts that exceed roughly 30 lines, handle repeated auth steps, or run parallel tasks across multiple targets. Instead of running commands, it writes the result to OUTPUT_DIR/artifacts/<task_name>.<ext>, checks syntax, and returns the dependencies alongside an execution command.
It applies concrete patterns: concurrent.futures.ThreadPoolExecutor for multiple hosts, requests.Session when making more than 3 HTTP requests to one host, and timeouts on every I/O call. Broad except: blocks are barred, while secrets stay isolated as variables at the top of the file rather than hardcoded inline.
Execution remains entirely manual, so reviewing the output and running the final file is your responsibility.
What it helps with
- generate parallel Python scripts with ThreadPoolExecutor
- write Bash or PowerShell automation scripts exceeding 30 lines
- set up scripts using impacket or ldap3 with strict I/O timeouts
- validate script syntax prior to manual execution
Example prompts for the agent
- Generate a python3 script to query hostnames concurrently using impacket.
- Create a bash script to collect logs from multiple IPs with strict timeouts.
- Write a powershell task that reuses credentials at the top of the file.
How to install script-generator
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill script-generator -a claude-code2Manually: put the files in place
~/.claude/skills/script-generator/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/script-generator/
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill script-generator -a codex2Manually: put the files in place
.agents/skills/script-generator/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill script-generator -a cursor2Manually: put the files in place
.cursor/skills/script-generator/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install transilienceai/communitytools skills/script-generator2Manually: put the files in place
.github/skills/script-generator/SKILL.md1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill script-generator -a gemini-cli2Platform CLI
gemini skills install https://github.com/transilienceai/communitytools --path skills/script-generatorManually: put the files in place
.devin/skills/script-generator/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill script-generator -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about script-generator
How to install script-generator in Claude Code?
Run npx skills add transilienceai/communitytools --skill script-generator -a claude-code in your terminal. This creates ~/.claude/skills/script-generator/SKILL.md.
Does script-generator run the code it produces?
No, the skill is restricted to generating, optimizing, and validating syntax. It never runs scripts.
What languages are supported by script-generator?
It accepts requests for python3, powershell, and bash.
Source text
SKILL.md · MIT
--- name: script-generator description: Generates optimized, syntax-validated scripts on demand. Never executes — only generates, optimizes, and validates. --- # Script Generator Generates optimized, syntax-validated scripts on demand. **Never executes scripts.** ## When to Use - Scripts exceed ~30 lines - Parallel operations on multiple targets - Multi-library patterns (impacket + ldap3, pypsrp + concurrent.futures) - Repeated auth handshakes or connection setup ## Request Format ``` LANGUAGE: python3 | powershell | bash TASK: What the script should accomplish TARGETS: IPs, hostnames, URLs CREDENTIALS: user, pass, hash, domain, certs AVAILABLE_LIBRARIES: What's installed OUTPUT_FORMAT: stdout format, file writes CONSTRAINTS: timeout, no destructive ops, output directory CONTEXT: (optional) Prior output, errors, what failed ``` ## Optimization - Multiple targets → `concurrent.futures.ThreadPoolExecutor` - >3 HTTP requests to same host → `requests.Session` - Repeated auth → single auth, reuse session/token - Prefer high-level libraries (impacket, ldap3, requests) ## Output Write to `OUTPUT_DIR/artifacts/<task_name>.<ext>`. Return: ``` SCRIPT_PATH: OUTPUT_DIR/artifacts/task_name.py LANGUAGE: python3 VALIDATION: PASSED EXECUTION: python3 OUTPUT_DIR/artifacts/task_name.py DEPENDENCIES: impacket, concurrent.futures (stdlib) ``` ## Rules - Never execute scripts — only generate, optimize, validate - Per-operation error handling — no bare `except:` - Timeout enforcement on all I/O - Validate syntax before returning - No secrets hardcoded — credentials as variables at top
© transilienceai · MIT · transilienceai/communitytools