Open Paws fork of peteromallet/desloppify. Adds advocacy-specific detectors for speciesist language and activist security antipatterns, persona-based browser QA, and Windows platform fixes — while tracking upstream for general improvements.
A multi-language codebase health scanner that combines mechanical detection (dead code, duplication, complexity, security) with subjective LLM review (naming, abstractions, module boundaries), then works through a prioritized fix loop. State persists across scans so it chips away over multiple sessions, and the scoring resists gaming — the only way up is genuinely better code.
29 languages supported. Full plugin depth for TypeScript, Python, C#, C++, Dart, GDScript, Go, and Rust. Generic linter + tree-sitter support for Ruby, Java, Kotlin, and 18 more.
Overall score = 25% mechanical + 75% subjective. A score above 98 should correlate with a codebase a seasoned engineer would call beautiful.
Detects speciesist language patterns in code, comments, and documentation across all 29 supported languages plus .md, .txt, .rst files. Rules are defined in YAML, sourced from project-compassionate-code.
| Category | Count | Examples |
|---|---|---|
| Idioms | 30 | "kill two birds with one stone", "beat a dead horse" |
| Metaphors | 21 | "sacred cow", "cash cow", "sacrificial lamb" |
| Insults | 6 | "code monkey", "cowboy coding" |
| Process language | 5 | "nuke", "cull", "kill process" |
| Terminology | 3 | "master/slave", "whitelist/blacklist", "grandfathered" |
Each finding includes a suggested replacement. Context suppression reduces false positives for technical terms (POSIX kill(), git master branch), proper nouns, and quotations.
Heuristic detector for activist protection antipatterns based on three adversaries:
- State surveillance — ag-gag statutes, warrants, device seizure
- Industry infiltration — corporate investigators, social engineering
- AI model bias — training data encoding speciesist defaults, telemetry leakage
Detects: identity leakage in logs/errors, sensitive data to external AI APIs without zero-retention headers, investigation materials in public paths, unencrypted writes of sensitive data, IP address logging, sensitive data in browser storage.
New persona-qa command for browser-based testing with configurable persona profiles (YAML). Findings integrate into the standard work queue alongside mechanical and subjective issues.
desloppify persona-qa --prepare --url https://example.com # generate agent instructions
# agent runs Playwright, captures findings in JSON
desloppify persona-qa --import findings.json # merge into state
desloppify persona-qa --status # per-persona summary
desloppify next # persona QA items appear in queueinput()blocking in TypeScript logs detector — replaced withisatty()guardmsvcrt.locking()infinite wait — replaced with 5s retry timeout (LK_NBLCK)- Dataclass JSON serialization crash on state save — fixed with
dataclasses.asdict()fallback
Three new mechanical dimensions (weight 1.0 each):
| Dimension | What it measures |
|---|---|
| Advocacy language | Speciesist language in code and docs |
| Advocacy security | Activist protection antipatterns |
| Persona QA | Browser-based usability findings |
Requires Python 3.11+.
pip install "git+https://github.com/Open-Paws/desloppify.git#egg=desloppify[full]"git clone https://github.com/Open-Paws/desloppify.git
cd desloppify
pip install -e ".[full]"uvx --from "git+https://github.com/Open-Paws/desloppify.git" desloppify scan --path .command -v desloppify >/dev/null 2>&1 && echo "desloppify: installed" || echo "NOT INSTALLED"The [full] extra includes tree-sitter, PyYAML (for advocacy rules), bandit, Pillow, and defusedxml. For minimal install, drop [full] — advocacy detectors will fall back to a built-in YAML parser.
Paste this into your agent's conversation or system prompt:
I want you to improve the quality of this codebase. Install and run desloppify.
Run ALL of the following (requires Python 3.11+):
pip install --upgrade "git+https://github.com/Open-Paws/desloppify.git#egg=desloppify[full]"
desloppify update-skill claude # installs the full workflow guide
# options: claude, cursor, codex, copilot, droid, windsurf, gemini
Add .desloppify/ to your .gitignore — it contains local state that shouldn't be committed.
Before scanning, exclude noise directories:
desloppify exclude node_modules
desloppify exclude dist
# exclude vendor, build output, generated code, worktrees, etc.
desloppify scan --path .
desloppify next
THE LOOP: run `next`. It tells you what to fix, which file, and the resolve command.
Fix it, resolve it, run `next` again. This is your main job.
Your goal is the highest possible strict score. The scoring resists gaming — the only
way to improve it is to actually make the code better.
Don't be lazy. Large refactors and small detailed fixes — do both with equal energy.
No task is too big or too small. Fix things properly, not minimally.
Use `plan` / `plan queue` to reorder priorities or cluster related issues. Rescan
periodically. The scan output includes agent instructions — follow them.
desloppify scan --path . # runs all mechanical detectors (including advocacy)
desloppify status # check scoresThe scan will tell you if subjective dimensions need review. Follow its instructions.
desloppify review --prepare # generates blind review packet
# launch subagents to score dimensions against the blind packet
desloppify review --import merged.json --scan-after-importSubjective scores start at 0% until reviewed. This is where most of the score lives.
desloppify next # shows top-priority execution item
desloppify plan # see the living plan
desloppify plan queue # compact execution queue view
desloppify plan reorder <pat> top # reorder priorities
desloppify plan cluster create <name> # group related issuesgit checkout -b desloppify/code-health
# The loop:
desloppify next # 1. get next item
# fix the code # 2. make the change
# run the resolve command next shows # 3. mark it done
git add <files> && git commit -m "..." # 4. commit when logical
desloppify plan commit-log record # 5. record the commit
desloppify next # 6. repeatWhen the queue is empty, go back to Phase 1. New issues will surface, cascades will resolve, priorities will shift.
| Command | Purpose |
|---|---|
scan --path . |
Analyze codebase, save findings |
status |
Summary of scores and dimension health |
next |
Top-priority item from execution queue |
backlog |
All open work items |
plan |
View/manage the living plan |
plan queue |
Compact execution queue |
plan reorder |
Reorder priorities |
plan cluster |
Group related issues |
resolve |
Mark issues fixed |
autofix |
Auto-fix (when available) |
review --prepare |
Generate blind review packet |
review --import |
Import review scores |
show <pattern> |
Display specific findings |
suppress |
Suppress false positives |
exclude <path> |
Exclude paths from scanning |
persona-qa |
Browser QA persona testing |
update-skill <agent> |
Generate agent skill file |
config |
Show/edit configuration |
directives |
Agent lifecycle directives |
viz / tree |
Dependency graph visualization |
langs |
Show supported languages |
When fixing issues from the advocacy detectors, follow this guidance.
Replace the flagged term with the first suggested alternative. Don't add comments explaining why. Don't wrap in feature flags. Just change it.
- Identity leakage: Remove PII from log/error output. Use pseudonymous identifiers. Never log IP, email, or real names in the same statement.
- External API calls: Add zero-retention headers or route through self-hosted inference.
- Public path exposure: Move investigation/evidence data out of
public/staticdirectories. Use encrypted storage.
Use these terms correctly in code, comments, and commit messages:
| Term | Meaning | NOT |
|---|---|---|
| Campaign | Coordinated advocacy action | Marketing campaign |
| Investigation | Covert evidence gathering | Bug investigation |
| Coalition | Cross-org partnership | Generic collaboration |
| Witness | Person providing testimony | Test witness |
| Sanctuary | Rescue facility for animals | Sandbox |
| Companion animal | Animal living with humans | Pet |
| Farmed animal | Animal in agriculture | Livestock |
This fork tracks peteromallet/desloppify as upstream. Fork-specific changes live in new files (advocacy detectors, persona QA command) and minimal patches to scoring constants and language configs. Upstream merges should be clean.
git fetch upstream
git merge upstream/mainIssues and PRs go to github.com/Open-Paws/desloppify.
For upstream bugs unrelated to the advocacy extensions, file at github.com/peteromallet/desloppify.
MIT — same as upstream.