scitex-dev

SciTeX Dev (scitex-dev)

---

Problem and Solution

#	Problem	Solution
1	~70 ecosystem packages drift apart -- versions, READMEs, sphinx setups, CLI conventions diverge faster than humans can audit	scitex-dev ecosystem audit-* -- audit-project, audit-cli, audit-mcp-tools, audit-python-apis, audit-skills, audit-summary enforce ~130 numbered rules across the whole ecosystem
2	Skills scattered across ~70 source repos -- AI agents can't discover them; humans can't review them	scitex-dev skills list / get / export -- aggregate every package's _skills/ and symlink them into ~/.claude/skills/scitex/ for live-edit dev loops
3	Coordinated releases need ten manual steps -- bump version, push tag, watch CI, verify PyPI, deploy — done per-package, multiplied by 70	scitex-dev ecosystem sync + validate-versions + gui open -- one-shot install/sync across hosts, version-mismatch detection, and a web dashboard with the live state
4	Bulk renames across 70 repos break cross-references -- import paths, doc references, symlinks — sed -i corrupts something every time	scitex-dev rename-symbols -- atomic rename with cross-reference updates, regex support, dry-run preview, git-safety guards
5	Lint rules drift from the API they enforce -- a renamed function in figrecipe leaves the rule pointing at a nonexistent symbol, and the bug only shows up months later	scitex-dev linter (engine, formerly scitex-linter) -- aggregates per-package rule plugins via the scitex_dev.linter.plugins entry point. Rules ship in the package whose API they enforce, so rename + rule + test land in one PR. Doc-block linting for .py / .ipynb / .md / .rst. scitex-dev linter sweep walks every package's README + docs in one shot

Installation

There are exactly two installs. The CLI, the MCP server, icon rendering and skills all ship in the base install — there is no per-feature menu to assemble, and nothing to forget to add.

# Everything you need to USE scitex-dev:
pip install scitex-dev

# Plus the maintainer toolchain (pytest, sphinx, ...) — for working ON it:
pip install scitex-dev[all]

[all] adds development and documentation tooling on top of the base install; it does not unlock features. Extra names are restricted to {all, dev, docs} by PS-225, because a per-feature extra is something somebody eventually pins — and a pin naming an extra that does not exist installs anyway, warns into a log nobody reads, and silently omits the capability.

Configuration

Copy .env.example to .env (gitignored) at your project root, then edit. CLI flags always override env vars. The full list (with inline comments) lives in .env.example.

scitex-dev reads optional config + cache from the canonical SciTeX local-state locations:

Path	Scope	Purpose
~/.scitex/dev/	user-global	per-user config, cache
<proj-root>/.scitex/dev/	project-local	overrides for the current repo

Project-local wins when both exist. Both are optional.

Architecture

scitex_dev/
├── _cli/
│   ├── audit/                ← rule corpus (PA*, PS*, SK*, §*)
│   │   ├── _api/             ← Python-API rules (PA-1xx)
│   │   ├── _project/         ← project-structure rules (PS-1xx, PS-5xx)
│   │   ├── _skills/          ← skill-file rules (SK-1xx)
│   │   └── _summary/         ← CLI/MCP §-rules + audit-all wrapper
│   ├── ecosystem/            ← cross-package commands (audit-all, list, …)
│   └── _skills.py            ← `scitex-dev skills` group
├── _ecosystem/               ← shared helpers (skill-quality, ECOSYSTEM map)
├── _skills/                  ← canonical skill-file corpus shipped to agents
└── testing/                  ← `audit_all_for_package` pytest helper

Audit rules live alongside the code they check. Each rule has a docstring, an entry in RULES, a severity in _SEVERITY_OVERRIDES, and at least one unit test under tests/.../_audit/_project/.

Demo

flowchart LR
    A["scitex-dev ecosystem<br/>audit-all <pkg>"] --> B[audit-cli]
    A --> C[audit-mcp-tools]
    A --> D[audit-skills]
    A --> E[audit-python-apis]
    A --> F[audit-project]
    B & C & D & E & F --> G{any error?}
    G -- "yes" --> H["exit 1<br/>(CI fails)"]
    G -- "no" --> I["exit 0<br/>(CI green)"]

scitex-dev is a CLI/audit tool — its "demo" is the audit running on a package. Sample output (run on scitex-io):

=== audit-cli ===
ok    scitex-io: no CLI convention violations

=== audit-mcp-tools ===
ok  scitex-io: no MCP convention violations

=== audit-skills ===
ok  scitex-io: no skills violations

=== audit-python-apis ===
ok  scitex-io: no Python API violations

=== audit-project ===
ok  scitex-io: no project-structure violations

Failure mode (sample, scitex-dsp before the Hilbert fix):

=== audit-project ===
fail  scitex-dsp (/home/ywatanabe/proj/scitex-dsp): 2 error(s)
  [E] [PS-141 §1] README.md: missing mandatory `## Demo` section
  [E] [PS-503 §5] examples/01_demo_out: no FINISHED_SUCCESS/<session_id>/ subdir

See examples/ for runnable demos that exercise scitex-dev's own commands (search, version-management, docs aggregation).

Four Interfaces

Version, CI, documentation and error-handling helpers are re-exported from scitex_dev (internals live under _release, _docs, _core; don't import those paths). The active-project selector lives in scitex_dev.project.

# Version management
from scitex_dev import detect_mismatches, fix_local, fix_remote, verify_versions
mismatches = detect_mismatches()
fix_local(packages=["scitex-stats"], confirm=True)
verify_versions()

# CI/CD
from scitex_dev import check_ci, get_failing_packages, check_pypi_publish
status = check_ci()
failing = get_failing_packages()

# Docs & search
from scitex_dev import get_docs, build_docs, search
get_docs(package="scitex-writer")
search("save figure")

# Deployment
from scitex_dev import deploy_scitex_hub, verify_production
deploy_scitex_hub(confirm=True)

# LLM-friendly types
from scitex_dev import Result, ErrorCode, supports_return_as

# Active project (scitex-dev >=0.62.0)
from scitex_dev.project import resolve_project
resolve_project(explicit="synthetic/owned")

Skills are managed through the CLI (scitex-dev skills list/get/export), not a Python module.

# Ecosystem management
scitex-dev ecosystem list
scitex-dev ecosystem list --versions
scitex-dev ecosystem fix-mismatches --dry-run
scitex-dev ecosystem sync

# Active project (scitex-dev >=0.62.0)
scitex-dev project use synthetic/owned
scitex-dev project get --json
scitex-dev project clear

# Documentation
scitex-dev docs --package scitex-writer
scitex-dev search-docs "save figure"

# Bulk rename
scitex-dev rename-symbols old_name new_name --dry-run

# See all commands
scitex-dev --help
scitex-dev --help-recursive

Project references are owner/name or all. Resolution checks an explicit Python argument first, then SCITEX_PROJECT, then the persisted selection under $SCITEX_DIR/scitex-dev/active_project (default ~/.scitex/scitex-dev/active_project). project current remains a deprecated alias for project get; its warning goes to stderr so JSON stays usable. Selecting a project does not authorize access to it or establish store tenancy.

# Start server
scitex-dev mcp start

# Check setup
scitex-dev mcp doctor
scitex-dev mcp list-tools

# Installation info
scitex-dev mcp installation

Claude Code Setup — add .mcp.json to your project root. Use SCITEX_DEV_ENV_SRC to load configuration from a .src file:

{
  "mcpServers": {
    "scitex-dev": {
      "command": "scitex-dev",
      "args": ["mcp", "start"],
      "env": {
        "SCITEX_DEV_ENV_SRC": "${SCITEX_DEV_ENV_SRC}"
      }
    }
  }
}

Switch environments via your shell profile:

# Local machine
export SCITEX_DEV_ENV_SRC=~/.scitex/dev/local.src

# Remote server
export SCITEX_DEV_ENV_SRC=~/.scitex/dev/remote.src

Skills provide workflow-oriented guides that AI agents query to discover capabilities and usage patterns.

scitex-dev skills list              # List available skill pages
scitex-dev skills get SKILL         # Show main skill page
scitex-dev skills export --package scitex-dev  # Export to Claude Code
# Private skills (~/.scitex/*/skills/*-private/) are symlinked automatically

Full skills directory

Skill	Content
result-types	Result, ErrorCode, @supports_return_as for LLM-friendly responses
cli-mcp-utils	CLI and MCP utility helpers
config	Package configuration and priority config patterns
versions	Version management, mismatch detection and fixing
ecosystem	Ecosystem list, sync, and commit workflows
rename	Safe bulk rename with cross-reference updates
docs-search	Documentation aggregation and unified search
test-runner	Local and HPC test execution
full-update / full-update-deploy	End-to-end audit → bump → release → sync pipelines
agentic-test-overview / -skills / -mcp	Agent-vs-agent quality testing
dynamic-audit	Cross-package periodic audit checklist
env-vars	SCITEX_DEV_* / SCITEX_* environment variable reference

Part of SciTeX

scitex-dev is part of SciTeX. Install via the umbrella with pip install scitex[dev] to use as scitex.dev (Python) or scitex dev ... (CLI).

SciTeX follows the Four Freedoms for Research below, inspired by the Free Software Definition:

Four Freedoms for Research

0. The freedom to run your research anywhere — your machine, your terms.
1. The freedom to study how every step works — from raw data to final manuscript.
2. The freedom to redistribute your workflows, not just your papers.
3. The freedom to modify any module and share improvements with the community.

AGPL-3.0 — because we believe research infrastructure deserves the same freedoms as the software it runs on.

---