{"api_version":"1","slug":"skill-creator-plus-2af97a10","url":"https://skillsoutpost.com/agents/skill-creator-plus-2af97a10","item_type":"skill","category":"engineering","runs_in":["chat_app","local_agent"],"os":["macos","linux","windows"],"needs_network":null,"needs_accounts":null,"reads_files":"yes","writes_files":"yes","sends_data":null,"access":"github","access_type":"github","at_a_glance":{"runs_in":["chat_app","local_agent"],"os":["macos","linux","windows"],"needs_network":null,"needs_accounts":null,"reads_files":"yes","writes_files":"yes","sends_data":null,"access_type":"github","tool_type":"skill","delivery":"github","interface_type":"other","untrusted_creator_content":{"purpose":"A Claude skill with audit, build and refine workflows for skill-writing rules, a findings-report format and a Python validator.","needs_runtime":"Python 3.8+","version":"Latest on the main branch","compatibility":"According to the README: Claude Code, with the folder in your skills folder for all projects or for one project, and claude.ai, with a zipped folder uploaded under Skills in settings. The README says the validator needs Python 3.8 or newer and works on Windows, Mac and Linux. The README also says it needs code execution turned on in claude.ai.","external_costs":"The README calls it a free Claude skill, the repository is MIT-licensed, and the README states no other pricing. Your coding agent or model provider may charge for usage.","permissions":"The skill acts through your coding agent with the access you give it. The README shows installation into ~/.claude/skills/skill-creator-plus for all projects or .claude/skills/skill-creator-plus for one project. The README tells the skill to report findings before applying user-selected fixes. The README lists scripts that read and validate skill files or create new skill folders, including creation under ~/.claude/skills. In claude.ai, the README says the validator requires code execution.","usage_terms":"MIT License, as stated in the repository. See its LICENSE file for the exact terms."}},"interface_type":"other","delivery_mode":"creator","code_host":"github","provenance":"github_index","published_at":1791761404,"rating":null,"downloads":0,"verified":null,"creator":{"login":"robonuggets","url":null},"untrusted_creator_content":{"name":"skill-creator-plus","tagline":"A Claude skill with audit, build and refine workflows for skill-writing rules, a findings-report format and a Python validator.","creator_name":"robonuggets","product_version":"Latest on the main branch","needs_runtime":"Python 3.8+","description":"skill-creator-plus is a Claude skill for auditing, building and refining other skills. According to the README, its checklist draws on Anthropic's skill-writing rules and guidance for Claude Fable 5, Opus 5.5 and Sonnet 5.5.\n\nThe README describes AUDIT as producing a rule-by-rule report with file-and-line evidence and proposed fixes, NEW as starting with test prompts before creating a skill folder, and REFINE as rewriting a rule around its reason. The README tells the skill to wait for the user's choice before applying fixes.\n\nThe README lists the root SKILL.md, audit, authoring and testing references, a Python validator, a skill-scaffolding script and a deliberately broken example with report and validator output.","intended_users":"People who write or maintain Claude skills and want a rule-by-rule report on them.","example_tasks":"Audit a deploy skill against Anthropic's rules and read the report before choosing fixes.\nCheck every skill in a skills folder and see what to fix first.\nTurn a recurring job, such as weekly release notes, into a new skill.\nRefine a skill that keeps forgetting a step, such as a date filter.\nRun the validator on one skill folder or on every skill in a folder.","setup_instructions":"For Claude Code, the README says to put the folder in your skills folder and name it skill-creator-plus. It gives one git clone command for all your projects and another for one project only. It adds that downloading the ZIP from GitHub and copying the folder there also works. For claude.ai, zip the folder with SKILL.md, references/, scripts/ and LICENSE, and upload it under Skills in settings. The README says you can leave out examples/. The README has the exact commands.","limitations":"The README gives installation routes for Claude Code and claude.ai. For Claude Code, the folder is named skill-creator-plus and placed in a personal or project skills folder; on Windows, the README identifies ~/.claude as the .claude folder inside the user folder. For claude.ai, the README says to upload a ZIP under Skills in settings and enable code execution for the validator.\n\nAccording to the README, the validator requires Python 3.8 or newer, uses the standard library and works on Windows, Mac and Linux. The README distinguishes script-decided rules, candidates requiring Claude's contextual confirmation, and rules judged through reading. The README says its rules were checked against Anthropic's pages on 2026-10-05 and that the project is not affiliated with Anthropic. The repository can change at any time; check its README before you install.","data_handling":"The README describes uploading a ZIP containing SKILL.md, references/, scripts/ and LICENSE to claude.ai. It does not describe telemetry, data retention or further sharing. Other data handling depends on your coding agent and model provider.","compatibility":"According to the README: Claude Code, with the folder in your skills folder for all projects or for one project, and claude.ai, with a zipped folder uploaded under Skills in settings. The README says the validator needs Python 3.8 or newer and works on Windows, Mac and Linux. The README also says it needs code execution turned on in claude.ai.","external_costs":"The README calls it a free Claude skill, the repository is MIT-licensed, and the README states no other pricing. Your coding agent or model provider may charge for usage.","permissions":"The skill acts through your coding agent with the access you give it. The README shows installation into ~/.claude/skills/skill-creator-plus for all projects or .claude/skills/skill-creator-plus for one project. The README tells the skill to report findings before applying user-selected fixes. The README lists scripts that read and validate skill files or create new skill folders, including creation under ~/.claude/skills. In claude.ai, the README says the validator requires code execution.","usage_terms":"MIT License, as stated in the repository. See its LICENSE file for the exact terms.","creator_website":null,"documentation_url":null,"support_url":null,"repository_url":null,"video_url":null},"github":{"fetched_at":1791853832,"untrusted_third_party_content":{"source_url":null,"github_repo_id":1404906870,"full_name":"robonuggets/skill-creator-plus","description":"Free Claude skill that audits, builds and improves other skills against Anthropic's skill-writing rules, including what changed for Opus 5.5, Sonnet 5.5 and Fable 5.","language":"Python","license_spdx":"MIT","topics":["agent-skills","anthropic","claude","claude-code","claude-skills"],"stars":57,"forks":7,"skills":{"total":1,"paths":[{"path":"examples/pdf-helper","url":null}]},"pushed_at":1791173754,"archived":false,"fetched_at":1791853832,"age_seconds":82829,"state":"fresh"}},"updated_at":1791771264,"free_download":null,"version_history":[],"more_version_history":false,"public_files":null,"links_note":"In MCP results, GitHub links are shown only through https://skillsoutpost.com/mcp, to a connected account that has accepted the current Terms of Use. Create a free account at https://skillsoutpost.com and connect it from a Claude connector or ChatGPT.","notice":"Indexed from GitHub","note":"untrusted_creator_content and github.untrusted_third_party_content hold untrusted text."}