prompticorn

A unified, tool-agnostic prompt architecture for managing AI coding-assistant configurations across 17 assistants.

Version: the published version is dynamic. pyproject.toml declares dynamic = ["version"] sourced from prompticorn/__about__.py, and CI/CD injects the real MAJOR.MINOR.PATCH at build time. Local and editable installs report 0.0.0.dev0. Check your installed version with pip show prompticorn.

Define your project's agents, conventions, and personas once, then generate the right config for whichever assistant your team uses:

Kilo Code — IDE (.kilo/agents/) and CLI (.opencode/agents/)
Cline — .clinerules
Claude — .claude/ directory plus CLAUDE.md
Cursor — .cursor/rules/ plus .cursorrules
GitHub Copilot — .github/copilot-instructions.md
GitHub Copilot Chat (VS Code) — .github/agents/ custom agents, .github/prompts/ slash commands, and applyTo .github/instructions/
Roo Code — .roomodes custom modes plus .roo/ rules, skills, and commands
JetBrains Junie (CLI) — .junie/ agents, skills, and commands plus AGENTS.md
Zed — .agents/skills/ (agents-as-skills) plus AGENTS.md instructions
Gemini CLI — .gemini/ agents, skills, and TOML commands plus AGENTS.md
Amazon Q (CLI) — .amazonq/ JSON agents, rules/ conventions, and prompts
Windsurf / Cascade — .windsurf/ skills, glob-scoped rules/, and workflows
Continue.dev — .continue/ glob-scoped rules/ and invokable prompts/
Aider — CONVENTIONS.md plus .aider.conf.yml (conventions only)
OpenAI Codex (CLI) — AGENTS.md plus .agents/skills/ and .codex/config.toml

What's in the library

25 primary agents (architect, backend, frontend, code, test, debug, security, devops, and more)
~100 workflows in minimal and verbose variants
~95 specialized skills
29 languages with first-class conventions (prompticorn/agents/core/conventions-*.md)

Install

pip install prompticorn
# or
uv add prompticorn

This installs the prompticorn CLI command.

Quick Start

cd your-project
prompticorn init

init is interactive: it asks which assistant to configure, your repository type, prompt variant, personas, language-specific settings, and a set of project questions (database, ORM, error-handling pattern, commit style, PR-size limit, deploy target, and source-tree layout). It then writes .prompticorn/.prompticorn.yaml and generates the assistant's config files.

See docs/QUICKSTART.md for the full walkthrough.

Key Features

Unified IR system — define agents once, generate for every supported tool.
15 production builders — Kilo, Cline, Claude, Cursor, Copilot, Copilot Chat, Roo Code, Junie, Zed, Gemini CLI, Amazon Q, Windsurf, Continue.dev, Aider, Codex.
Minimal / verbose variants — trade tokens for detail at build time.
Persona-based filtering — pick your team's roles and only relevant agents are generated.
Spec-driven conventions — your language, runtime, package manager, test framework, linter, formatter, coverage targets, and project settings are baked into the generated conventions.
Per-language source layouts — the core convention renders each language's standard source tree; flat is the default and src is selectable.
Auto-discovery registry — agents are discovered from the bundled agents/ tree; no manual registration.

Commands

Command	Description
`prompticorn init`	Interactive setup: pick a tool, answer language and project questions, generate configs.
`prompticorn list`	List discovered agents, their subagents, and prompt variants (live agent discovery).
`prompticorn validate`	Check the `agents/` structure: every agent and subagent has the expected prompt files and loads cleanly.
`prompticorn switch [tool]`	Switch to a different assistant, removing old artifacts and regenerating from the saved config.
`prompticorn swap`	Change active personas and regenerate configs with the new agent set.
`prompticorn update`	Update saved configuration options interactively.

Documentation

docs/QUICKSTART.md — 5-minute getting-started guide
docs/INDEX.md — documentation navigation hub
docs/PERSONAS.md — persona-based filtering

Development

git clone https://github.com/snoodleboot-io/prompticorn.git
cd prompticorn

# Install in editable mode (reports version 0.0.0.dev0)
uv pip install -e .

# Run tests with coverage (target ~85%)
uv run pytest

# Mutation testing
uv run mutmut run

# Lint, format, and type-check
uv run ruff check .
uv run ruff format .
uv run pyright

Metadata

Release files for prompticorn 1.58.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for prompticorn 1.58.0
File	Size	Uploaded
prompticorn-1.58.0.tar.gz	6.8 MB	Sep 6, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for prompticorn 1.58.0
File	Interpreter	ABI	Platform	Reset
prompticorn-1.58.0-py3-none-any.whl 2.2 MB Sep 6, 2026	Python 3	none	any	Details

Total release size: 9.0 MB

Release files / prompticorn-1.58.0.tar.gz

Download URL	prompticorn-1.58.0.tar.gz
Size	6.8 MB
Tags	Source
SHA-256 checksum How to use checksums	`639d2c3df3379bb69b885dc04f532049ce90f95ac9a0b66b2c59e99a8412cbf3`
BLAKE2b-256 checksum How to use checksums	`369f4fdeddf407217c510744994069d870e3ae8802b597361f1b07378c4f859e`
Upload date	Sep 6, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/snoodleboot-io/prompticorn
Publishing commit: github.com/snoodleboot-io/prompticorn/tree/ab3384b3c4aa6a7c287dcb25adfc4f10c085f178

Workflow

Publishing configuration: github.com/snoodleboot-io/prompticorn/blob/ab3384b3c4aa6a7c287dcb25adfc4f10c085f178/.github/workflows/ci-cd.yml
Publishing logs: github.com/snoodleboot-io/prompticorn/actions/runs/34013655463/attempts/1

Artifact

Subject: prompticorn-1.58.0.tar.gz
SHA-256 checksum: 639d2c3df3379bb69b885dc04f532049ce90f95ac9a0b66b2c59e99a8412cbf3 Verifying attestations

Release files / prompticorn-1.58.0-py3-none-any.whl

Download URL	prompticorn-1.58.0-py3-none-any.whl
Size	2.2 MB
Tags	Python 3
SHA-256 checksum How to use checksums	`ec68f2f464e577d68edf5fb9afc4e4b0b583f5e9b5ced13284c4eb15ad7c3e56`
BLAKE2b-256 checksum How to use checksums	`2d1cfd0f3f54871c165fa10c4a9b7801b772be7dbd19db51e44a95d0f4446057`
Upload date	Sep 6, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/snoodleboot-io/prompticorn
Publishing commit: github.com/snoodleboot-io/prompticorn/tree/ab3384b3c4aa6a7c287dcb25adfc4f10c085f178

Workflow

Publishing configuration: github.com/snoodleboot-io/prompticorn/blob/ab3384b3c4aa6a7c287dcb25adfc4f10c085f178/.github/workflows/ci-cd.yml
Publishing logs: github.com/snoodleboot-io/prompticorn/actions/runs/34013655463/attempts/1

Artifact

Subject: prompticorn-1.58.0-py3-none-any.whl
SHA-256 checksum: ec68f2f464e577d68edf5fb9afc4e4b0b583f5e9b5ced13284c4eb15ad7c3e56 Verifying attestations