Skip to content
Back to search

prompt-fixer

SkillWorks with Claude CodeGitHub

Resolve intent, compile context-aware prompts and direct builders toward checked delivery with less wasted context, questioning and rework.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/carlosantano--prompt-fixer/badge.svg)](/listings/carlosantano--prompt-fixer)

HTML

<a href="/listings/carlosantano--prompt-fixer"><img src="/api/listings/carlosantano--prompt-fixer/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Clone, then install into **one existing project**. Python 3.12+ is needed only for the optional offline helper; no pip install is required.

git clone https://github.com/CarloSantano/prompt-fixer.git
cd prompt-fixer
python -B -m prompt_fixer install --host codex --project "/absolute/path/to/your-project" --dry-run
python -B -m prompt_fixer install --host codex --project "/absolute/path/to/your-project"
python -B -m prompt_fixer doctor --host codex --project "/absolute/path/to/your-project"

Replace the example path with your real project. On Windows use a quoted path such as `"C:\Users\you\Projects\my-app"`; `py -3.12` may replace `python` if that is how Python is installed. For Claude Code replace `--host codex` with `--host claude` in each command.

| Host | Installed project folder | Invocation after discovery | | --- | --- | --- | | Codex | `.agents/skills/prompt-fixer/` | `$prompt-fixer` plus your request | | Claude Code | `.claude/skills/prompt-fixer/` | `/prompt-fixer` plus your request |

Open or reload your project in the host and confirm `prompt-fixer` is in its skills list. Try:

$prompt-fixer build me an internal sales tool; choose the technical approach.

Use `/prompt-fixer` in Claude Code. Success means a visible comparison appears without building the underlying app. Say "execute that prompt" to proceed.

The installer refuses unmanaged or user-edited installs and changes no global settings. `doctor` verifies files and receipt, not discovery by a running host.

**Without Python:** copy the entire `skills/prompt-fixer/` folder, including its references, into the host's project folder above. Manual copying creates no installer receipt; manage later updates yourself.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

plugin.json · Manifest · 29 lines
  1. {
  2. "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
  3. "author": {
  4. "name": "Carlo Santo v3.0"
  5. },
  6. "description": "Resolve intent, compile context-aware prompts and direct builders toward checked delivery with less wasted context, questioning and rework.",
  7. "extensions": {
  8. "com.openai": {
  9. "interface": {
  10. "capabilities": [],
  11. "category": "Productivity",
  12. "defaultPrompt": [
  13. "Paversk mano laisvai parašytą mintį tikslingu darbu: parink vaidmenį, kontekstą, klausimus ir sėkmės patikrą.",
  14. "Padėk išgryninti, ką iš tikrųjų noriu sukurti, ir pradėk nuo svarbiausio žingsnio.",
  15. "Peržiūrėk šią užduotį: ar teisingas tikslas, darbo būdas ir rezultato patikra?"
  16. ],
  17. "developerName": "Carlo Santo v3.0",
  18. "displayName": "Prompt Fixer",
  19. "longDescription": "Prompt Fixer padeda laisvai parašytą mintį paversti tikslingu promptu arba vykdomu darbu. Pirmiausia atskiria klausimą, prompto perrašymą ir vykdymą. Išsaugo kontekstą, reikalavimus ir ambiciją, parenka naudingą atsakomybę bei metodus. Vibecoderiams padeda rinktis tinkamą techninį kelią, pristatyti naudojamą rezultatą ir sustoti pasiekus sutartą kokybę. Taupo kontekstą ir dėmesį vengdamas kartojimosi; faktinės tokenų ar pajamų naudos nežada be matavimo. Pritaikomas verslo ir asmeninėms užduotims. Aiškus @Prompt Fixer su padrika užduotimi pateikia pagerintą promptą; vykdo paprašius. Automatinis parinkimas priklauso nuo aplinkos.",
  20. "shortDescription": "Nuo minties iki aiškaus darbo"
  21. }
  22. }
  23. },
  24. "name": "prompt-fixer",
  25. "version": "0.7.0",
  26. "repository": "https://github.com/CarloSantano/prompt-fixer",
  27. "homepage": "https://github.com/CarloSantano/prompt-fixer#readme",
  28. "license": "MIT"
  29. }
README.md · README · 150 lines
  1. # Prompt Fixer
  2. [![Checks](https://github.com/CarloSantano/prompt-fixer/actions/workflows/ci.yml/badge.svg)](https://github.com/CarloSantano/prompt-fixer/actions/workflows/ci.yml)
  3. [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
  4. **Rough idea → visible prompt improvement → execute when you ask.**
  5. An instruction layer for your existing AI assistant. For builders who want to
  6. ship useful products, services and automations without learning prompt syntax.
  7. No extra server, API key or model call from this package. Your AI host's access,
  8. tools, permissions and usage costs still apply.
  9. [Lietuviškai: pradžia draugui](docs/QUICKSTART_LT.md) ·
  10. [Download the release](https://github.com/CarloSantano/prompt-fixer/releases/latest) ·
  11. [Evidence and limitations](qa/README.md)
  12. ## What you see
  13. 1. **Your request** — your actual words; long inputs use a marked excerpt.
  14. 2. **Improved prompt** — separately copyable instructions in your language.
  15. 3. **What changed** — one short account of concrete changes.
  16. A short status identifies rewriting versus execution. Ask for "only the final
  17. prompt" to skip the comparison. Specific user instructions remain in control.
  18. | Request | Behavior |
  19. | --- | --- |
  20. | Invoke with "build me a client-request app; choose the technology" | Show an improved prompt and stop before building. |
  21. | "Improve and execute now: add a CSV export to this project" | Inspect the available project and perform authorized work. |
  22. | "Does Prompt Fixer need a backend?" | Answer the question directly. |
  23. | Invoke without a task | Ask for the task. |
  24. Ordinary implicit use preserves the original mode. Automatic selection is
  25. host-dependent. This skill is not an interceptor or cross-chat memory service.
  26. ## Quick install: Codex or Claude Code
  27. Clone, then install into **one existing project**. Python 3.12+ is needed only
  28. for the optional offline helper; no pip install is required.
  29. ```sh
  30. git clone https://github.com/CarloSantano/prompt-fixer.git
  31. cd prompt-fixer
  32. python -B -m prompt_fixer install --host codex --project "/absolute/path/to/your-project" --dry-run
  33. python -B -m prompt_fixer install --host codex --project "/absolute/path/to/your-project"
  34. python -B -m prompt_fixer doctor --host codex --project "/absolute/path/to/your-project"
  35. ```
  36. Replace the example path with your real project. On Windows use a quoted path
  37. such as `"C:\Users\you\Projects\my-app"`; `py -3.12` may replace `python` if
  38. that is how Python is installed. For Claude Code replace `--host codex` with
  39. `--host claude` in each command.
  40. | Host | Installed project folder | Invocation after discovery |
  41. | --- | --- | --- |
  42. | Codex | `.agents/skills/prompt-fixer/` | `$prompt-fixer` plus your request |
  43. | Claude Code | `.claude/skills/prompt-fixer/` | `/prompt-fixer` plus your request |
  44. Open or reload your project in the host and confirm `prompt-fixer` is in its
  45. skills list. Try:
  46. ```text
  47. $prompt-fixer build me an internal sales tool; choose the technical approach.
  48. ```
  49. Use `/prompt-fixer` in Claude Code. Success means a visible comparison appears
  50. without building the underlying app. Say "execute that prompt" to proceed.
  51. The installer refuses unmanaged or user-edited installs and changes no global
  52. settings. `doctor` verifies files and receipt, not discovery by a running host.
  53. **Without Python:** copy the entire `skills/prompt-fixer/` folder, including its
  54. references, into the host's project folder above. Manual copying creates no
  55. installer receipt; manage later updates yourself.
  56. ## ChatGPT Work Mode
  57. If your account supports creating custom plugins, download the portable ZIP
  58. from [Releases](https://github.com/CarloSantano/prompt-fixer/releases/latest),
  59. attach it in Work Mode with Plugin Creator available, and ask it to create a
  60. **private personal plugin from the attached archive**. Select your own Prompt
  61. Fixer plugin and mention it with your request.
  62. This depends on your account's features. An ordinary chat does not automatically
  63. install a repository. The author's private plugin grants no access to another
  64. account. This is a portable source distribution, not a marketplace listing.
  65. ## Update or remove
  66. Run `git pull --ff-only` in this source checkout, then rerun installation with
  67. `--replace`. An unchanged managed install is backed up in the target project's
  68. `.prompt-fixer/backups/` before replacement.
  69. ```sh
  70. python -B -m prompt_fixer install --host codex --project "/absolute/path/to/your-project" --replace
  71. python -B -m prompt_fixer uninstall --host codex --project "/absolute/path/to/your-project"
  72. ```
  73. Removal moves unchanged managed files to a recoverable backup. Reconcile user
  74. edits first; the helper refuses to overwrite or remove them. No auto-update or
  75. global configuration is added.
  76. ## Offline checks
  77. ```sh
  78. python -B -m unittest discover -s tests -v
  79. python -B -m prompt_fixer check --evidence
  80. python -B qa/evalkit.py validate
  81. python -B qa/downstream_checks.py candidate ingest
  82. python -B qa/downstream_checks.py candidate approval
  83. python -B tools/build_release.py
  84. ```
  85. These commands make no API/model calls. Run outputs stay outside the source
  86. bundle. CI checks Windows and Linux. Historical response replay is separate from
  87. correspondence to current recorded evidence; neither executes a new AI model.
  88. Instruction or helper-code changes invalidate current evidence.
  89. Evidence includes a small same-model study and targeted fresh-context checks.
  90. It does **not** establish human validation, universal activation, production
  91. integrations, billed-token savings or improved revenue.
  92. To compare actual caller-supplied host/API usage for the whole path:
  93. ```sh
  94. python -B -m prompt_fixer measure examples/usage-ledger.json
  95. ```
  96. Count fixer, execution and repair steps. The empty template reports unknown
  97. usage; missing counts are never zero. Successful outcomes, matched conditions
  98. and the same execution model are required. Cache/reasoning counts are subsets,
  99. not extra totals. The helper cannot capture telemetry, verify outcomes or detect
  100. omitted steps independently. Never publish private task data in a usage ledger.
  101. ## Design
  102. Resolve intent, preserve constraints and selectively load references for roles,
  103. questions, architecture, methods and completion. Distinguish assumptions from
  104. requirements; ask questions that change a meaningful decision. Personal work
  105. needs no business interview. Roles identify useful responsibilities, not
  106. fictional experience. Approvals retain the exact target/recipient and reviewed
  107. content/version. Prompt text alone does not enforce production security.
  108. Skill locations follow the [Codex documentation](https://learn.chatgpt.com/docs/build-skills)
  109. and [Claude Code documentation](https://code.claude.com/docs/en/skills), checked
  110. 2026-11-21. Verify actual discovery in your own session.
  111. See [contributing](CONTRIBUTING.md), [changes](CHANGELOG.md),
  112. [provenance](NOTICE.md), [public-distribution notes](docs/PUBLIC_DISTRIBUTION.md)
  113. and the [MIT license](LICENSE).

Version history

  • v0.7.011/22/2026

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.