Borrowing it
Nothing to install: this file belongs to drunkrhin0/antislop. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/drunkrhin0/antislop/main/AGENTS.mdgit clone --depth 1 https://github.com/drunkrhin0/antislopWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/drunkrhin0/antislop/agents-md)<a href="https://agentmods.dev/instructions/drunkrhin0/antislop/agents-md"><img src="https://agentmods.dev/badge/instructions/drunkrhin0/antislop/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/drunkrhin0/antislop/agents-md"><img src="https://agentmods.dev/badge/instructions/drunkrhin0/antislop/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.07180 | $0.07180 |
| Opus 5.5 | $0.02872 | $0.02872 |
| Sonnet 5.5 | $0.01436 | $0.01436 |
| Haiku 4.5 | $0.00718 | $0.00718 |
Grade B, and why
antislop AGENTS.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Enumerates other installed skillsmediumAgent snooping
Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.
# Test a skill locally — copy to agent skills dir cp skills/antislop/SKILL.md ~/.claude/skills/antislop/ How it starts
The opening of the file, as written. The whole thing — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS
Agent skills
Issue tracker
Specs and implementation tickets are tracked as Forgejo issues. See docs/agents/issue-tracker.md.
Protected Forgejo writes
Agent-authored Forgejo file mutations are branch-first and fail closed on
main, the resolved default branch, and named protected branches. Use
docs/agents/forgejo-write-policy.md and the public
tools/forgejo_write_policy.py contract before any mutation. Create and verify a
codex/ task branch, record its head, mutate only that branch through Forgejo
MCP via Bifrost, verify the head after each mutation batch, and open a PR.
Creating a PR never grants approval or merge authority.
Triage labels
Issues use the labels configured on this repo's Forgejo instance, not the default Matt Pocock set. See docs/agents/triage-labels.md for the live list.
Domain docs
This is a single-context repository with a root CONTEXT.md and system decisions under docs/adr/. See docs/agents/domain.md.
What this project does
Antislop ships two AI agent skills. antislop is an ambient writing style that triggers automatically when asked to write or edit prose. Its audit mode scores text 0-100 and returns a violations list with severity and excerpts. generate-slop creates user-invoked synthetic material for demos and evaluation fixtures and must never be selected by an agent on its own. Works with any agent supporting the SKILL.md or skills system.
There is also an agent file (.opencode/agents/antislop.md) that bundles both modes into a single spawnable agent. The agent file format is opencode-specific, but the rules content works as a system prompt in any LLM. generate-slop remains a separate first-class skill across plugin surfaces because its explicit inverse purpose must not become a mode of the public antislop router.
How to run and test
# Full propagation check: generate.py --check + validate.py, one report
bash check.sh
# Unit tests (stdlib unittest, no external deps)
python3 -m unittest discover -s tests -t .
# Structural invariants only (version read from rules.json)
python3 tools/validate.py --skills-dir skills --expect-version-from rules.json
# Committed artifacts match the registry
python3 tools/generate.py --check
# Score arbitrary text against the registry
echo "some text" | python3 tools/score.py --profile general
# Fidelity gate (patina-style, issue #90)
python3 tools/fidelity.py --source-text "before" --candidate-text "after"
python3 tools/fidelity.py --source before.txt --candidate after.txt --authorize number=50
# Semantic drift fixtures (slopkit-style, issue #91)
python3 tools/drift.py
# Clarity-style review (issue #96)
python3 tools/review.py --source-text "..." --medium argument
python3 tools/review.py --fixtures skills/antislop/evals/clarity-review-fixtures.json
# Protected repair (unslop-style, issue #92)
python3 tools/repair.py --file doc.md
cat doc.md | python3 tools/repair.py --format json
python3 tools/repair.py --file doc.md --write --format diff
python3 tools/repair.py --fixtures skills/antislop/evals/repair-fixtures.json
# Bounded repair loop (lynote-style, issue #93)
python3 tools/repair_loop.py --file doc.md
cat doc.md | python3 tools/repair_loop.py
python3 tools/repair_loop.py --file doc.md --max-attempts 2 --char-budget 100
python3 tools/repair_loop.py --fixtures skills/antislop/evals/repair-loop-fixtures.json
# Sepia-style operation and venue routing (sepia-style, issue #94)
python3 tools/sepia.py review --source-text "..." --venue ticket
python3 tools/sepia.py review --source-text "..." --fiction
python3 tools/sepia.py refactor --source-text "..." --venue ticket --accept vocab-utilize=use
python3 tools/sepia.py recreate --source-text "..." --candidate "..." --venue postmortem
python3 tools/sepia.py --fixtures skills/antislop/evals/sepia-routing-fixtures.json
# Evidence-bound delivery envelope (review-write style, issue #95)
python3 tools/delivery.py verify --source-text "..." --body-text "..."
python3 tools/delivery.py envelope --source-text "..." --body-text "..." --medium argument
python3 tools/delivery.py check --envelope envelope.json
python3 tools/delivery.py --fixtures skills/antislop/evals/delivery-envelope-fixtures.json
# Say-It-Human evidence and locale routing (say-it-human style, issue #97)
python3 tools/say_human.py label --source-text "..." --claim "..."
python3 tools/say_human.py repair --source-text "..." --required-claim "..." --locale zh-CN --venue release-note --authorize-delete
python3 tools/say_human.py locale --source-text "..." --locale zh-CN
python3 tools/say_human.py --fixtures skills/antislop/evals/say-human-fixtures.json
# Tagore-style substance report (issue #102)
python3 tools/substance.py --source-text "..." --medium argument
cat doc.md | python3 tools/substance.py --medium reference
python3 tools/substance.py --proposal proposal.json
python3 tools/substance.py --fixtures skills/antislop/evals/substance-fixtures.json
# Staged scan (humanizer-stack style, issue #98)
python3 tools/scan.py --file doc.md
cat doc.md | python3 tools/scan.py
python3 tools/scan.py --doc a.md --doc b.md --doc c.md
python3 tools/scan.py --file doc.md --fail-on-advisory
python3 tools/scan.py --fixtures skills/antislop/evals/staged-scan-fixtures.json
# Lint via CI (requires Docker)
act push -W .forgejo/workflows/lint-skills.yml
# Test a skill locally — copy to agent skills dir
cp skills/antislop/SKILL.md ~/.claude/skills/antislop/
cp skills/generate-slop/SKILL.md ~/.claude/skills/generate-slop/
# Test the agent — copy to global agents dir
cp .opencode/agents/antislop.md ~/.config/opencode/agents/
# Test the Claude Code plugin
claude plugin validate .claude-plugin/plugin.json
claude --plugin-dir .
# Point Kiro at the Power: Powers panel -> Add Custom Power -> Import power from a folder -> absolute path to powers/antislop
# Kiro skills, manual path
cp -r skills/antislop skills/generate-slop ~/.kiro/skills/
# Verify the Kiro workspace exposes both skills
test -f .kiro/skills/antislop/SKILL.md || test -L .kiro/skills/antislop
test -f .kiro/skills/generate-slop/SKILL.md
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +84 lines · +3,129 tokens per session scan C → B 29f81058e8df
- 1mo ago First seen · 195 lines · 4,051 tokens per session scan C 2f01210f011b
antislop AGENTS.md is an instructions file published in the GitHub repository drunkrhin0/antislop (23 stars, last pushed 1mo ago), licensed MIT. It adds 7,180 tokens to every session, about $0.0287 per session on Opus 5.5. A static security scan graded it B with 1 finding (enumerates other installed skills). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.