obey-cli 0.1.0__tar.gz → 0.1.1__tar.gz

{obey_cli-0.1.0 → obey_cli-0.1.1}/CHANGELOG.md

@@ -3,6 +3,16 @@
 All notable changes to this project are documented here. The format follows Keep a Changelog, and the project uses
 semantic versioning.
 
+## [0.1.1] - 2026-10-24
+
+Documentation and packaging only; the CLI, the checkers and the hooks behave as in 0.1.0.
+
+- The README opens with what obey is for and how to try it, installs from PyPI (`uvx --from obey-cli obey demo`),
+  and uses absolute image and link URLs, so it also renders on PyPI.
+- A new social preview, built from the study's numbers by `eval/make_social_preview.py`.
+- Release assets are only the wheel and the sdist; study charts are written with stable element ids, so rebuilding
+  them changes no bytes.
+
 ## [0.1.0] - 2026-10-24
 
 First release.

{obey_cli-0.1.0 → obey_cli-0.1.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: obey-cli
-Version: 0.1.0
+Version: 0.1.1
 Summary: Unit tests for AGENTS.md: turns agent instructions into checkable rules, runs your coding agent on tempting tasks in a sandbox, measures per-rule compliance from what it actually did, and compiles the rules it ignores into hooks.
 Project-URL: Homepage, https://github.com/B0yko/obey
 Project-URL: Issues, https://github.com/B0yko/obey/issues
@@ -29,40 +29,55 @@ Requires-Dist: matplotlib>=3.8; extra == 'study'
 Requires-Dist: numpy>=1.26; extra == 'study'
 Description-Content-Type: text/markdown
 
+<p align="center">
+  <img alt="obey, unit tests for your AGENTS.md: a report card of the rules a coding agent followed, and a hook that fixed the one it broke" src="https://raw.githubusercontent.com/B0yko/obey/main/.github/social-preview.png" width="860">
+</p>
+
 # obey
 
-**Unit tests for AGENTS.md.** obey turns the rules in your agent instruction files into checks, runs your coding agent
-on small tempting tasks in a sandbox, measures which rules it actually follows from what it did, and compiles the ones it
-ignores into hooks for Claude Code, Codex CLI, Cursor, Gemini CLI, Copilot CLI and OpenCode.
+**You wrote rules for your coding agent. Does it follow them?**
+
+AGENTS.md, CLAUDE.md and Cursor rules tell coding agents how to work in your repository, but nothing checks that they
+do. obey tests every rule against what the agent actually did, and turns the rules it breaks into hooks it cannot skip.
 
 [![CI](https://github.com/B0yko/obey/actions/workflows/ci.yml/badge.svg)](https://github.com/B0yko/obey/actions/workflows/ci.yml)
 [![Release](https://img.shields.io/github/v/release/B0yko/obey)](https://github.com/B0yko/obey/releases)
-[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](LICENSE)
+[![PyPI](https://img.shields.io/pypi/v/obey-cli)](https://pypi.org/project/obey-cli/)
+[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](https://github.com/B0yko/obey/blob/main/LICENSE)
 
-**On tempting tasks, Claude Haiku 5.5 in Claude Code and OpenCode followed the rule under test in all 1,474 exercised
-runs where it was in AGENTS.md, against 81% without it (stratified difference +17.6 points [95% CI 4.9–34.1]); the
-phrasing (imperative, prohibition, or with a reason) made no measurable difference, since every phrasing was at
-100%. Hooks without the rule raised compliance from 34% to 100% for the rules agents broke, and in 0 of 123 blocked
-runs did the agent break the rule another way.** This is probe compliance, measured after correcting 136 checker
-false positives (94% before the correction); see [how it was measured](study/REPORT.md).
+1. **Check.** `obey init` reads your instruction files and turns rules such as "use pnpm", "never push" or "run the
+   tests before you finish" into automatic checks.
+2. **Test.** `obey run` gives your agent a small task per rule where breaking it is the easy way out, in a sandbox,
+   and scores what it did (commands, diffs, commits), not what it said. You get a report per rule.
+3. **Enforce.** `obey enforce` writes hooks for Claude Code, Codex CLI, OpenCode, Copilot CLI, Gemini CLI and Cursor,
+   plus a git pre-commit check, so a broken rule is blocked the moment the agent tries.
 
-<picture>
-  <img alt="obey demo replaying recorded runs: per-rule compliance and the before/after table with hooks" src="docs/images/demo.svg">
-</picture>
+**What we found** in 2,663 sandboxed runs of Claude Haiku 5.5 in Claude Code and OpenCode, on test tasks built to tempt
+the agent: it followed a rule every time the rule was in AGENTS.md (1,474 of 1,474 runs) and 81% of the time when it
+was not. How the rule was worded made no measurable difference. For the rules it broke without the rule written down,
+a hook alone raised compliance from 34% to 100%, and in 0 of 123 blocked runs did the agent find a way around it.
+[How it was measured, the caveats, and the 136 errors we found in our own checkers](#results).
+
+### Try it in one command
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey demo
+uvx --from obey-cli obey demo
 ```
 
-The demo needs no API key and no Docker: it replays agent runs recorded in the study and scores them again.
+No API key and no Docker: the demo replays agent runs recorded in the study and scores them again.
+
+<picture>
+  <img alt="obey demo replaying recorded runs: per-rule compliance and the before/after table with hooks" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/demo.svg">
+</picture>
 
 ## Quickstart
 
-In your repository (Python 3.10+, git, [uv](https://docs.astral.sh/uv/)):
+In your repository (Python 3.10+, git, [uv](https://docs.astral.sh/uv/)). The package on PyPI is `obey-cli`; to get a
+plain `obey` command, install it with `uv tool install obey-cli` or `pipx install obey-cli`.
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey init      # extract rules into obey.yaml and create probes
-uvx --from git+https://github.com/B0yko/obey obey lint      # no agent runs: checks rules, hashes, enforcement; writes the badge
+uvx --from obey-cli obey init      # extract rules into obey.yaml and create probes
+uvx --from obey-cli obey lint      # no agent runs: checks rules, hashes, enforcement; writes the badge
 ```
 
 Read `obey.yaml`: every rule has a checker, or a status that says why not (`uncheckable` for rules that need taste,
@@ -70,23 +85,23 @@ Read `obey.yaml`: every rule has a checker, or a status that says why not (`unch
 
 ```bash
 # enforce now, no runs needed: hooks for every agent plus a git pre-commit check
-uvx --from git+https://github.com/B0yko/obey obey enforce --dry-run
-uvx --from git+https://github.com/B0yko/obey obey enforce
+uvx --from obey-cli obey enforce --dry-run
+uvx --from obey-cli obey enforce
 
 # measure: needs Docker (or colima) and an OpenRouter key; the key never enters the sandbox
-uvx --from git+https://github.com/B0yko/obey obey doctor --build-image          # once: the sandbox image, a few GB
+uvx --from obey-cli obey doctor --build-image          # once: the sandbox image, a few GB
 export OPENROUTER_API_KEY=...
-uvx --from git+https://github.com/B0yko/obey obey run --estimate --model anthropic/claude-haiku-5.5
-uvx --from git+https://github.com/B0yko/obey obey run --model anthropic/claude-haiku-5.5 --budget 2
-uvx --from git+https://github.com/B0yko/obey obey enforce --from .obey/report.json   # hooks for the rules it broke
+uvx --from obey-cli obey run --estimate --model anthropic/claude-haiku-5.5
+uvx --from obey-cli obey run --model anthropic/claude-haiku-5.5 --budget 2
+uvx --from obey-cli obey enforce --from .obey/report.json   # hooks for the rules it broke
 cp .obey/report.json before.json
-uvx --from git+https://github.com/B0yko/obey obey run --compare before.json --model anthropic/claude-haiku-5.5
+uvx --from obey-cli obey run --compare before.json --model anthropic/claude-haiku-5.5
 ```
 
 ### Test your AGENTS.md in 10 minutes
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey run --model anthropic/claude-haiku-5.5 --trials 3 --baseline-trials 1 --budget 2
+uvx --from obey-cli obey run --model anthropic/claude-haiku-5.5 --trials 3 --baseline-trials 1 --budget 2
 ```
 
 On the demo repository (11 checkable rules with probes) this made 44 runs of Claude Code 2.1.309 in 9 min 20 s and
@@ -151,12 +166,12 @@ from the vendor's documentation. Every target also gets the git pre-commit
 check. Codex CLI loads project hooks only after you trust them (`/hooks`); Copilot CLI's prompt mode needs
 `--trust-folder` or `GITHUB_COPILOT_PROMPT_MODE_REPO_HOOKS=true`.
 
-Details, flags and isolation checks: [docs/agents.md](docs/agents.md). Checker definitions:
-[docs/checkers.md](docs/checkers.md). Enforcement targets: [docs/enforce.md](docs/enforce.md).
+Details, flags and isolation checks: [docs/agents.md](https://github.com/B0yko/obey/blob/main/docs/agents.md). Checker definitions:
+[docs/checkers.md](https://github.com/B0yko/obey/blob/main/docs/checkers.md). Enforcement targets: [docs/enforce.md](https://github.com/B0yko/obey/blob/main/docs/enforce.md).
 
 ## Configuration
 
-`obey.yaml` is the reviewed list of rules ([reference](docs/config.md)). Environment variables:
+`obey.yaml` is the reviewed list of rules ([reference](https://github.com/B0yko/obey/blob/main/docs/config.md)). Environment variables:
 
 | Variable | Used by | Meaning |
 |---|---|---|
@@ -176,7 +191,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
-      - uses: B0yko/obey@v0.1.0
+      - uses: B0yko/obey@v0.1.1
 ```
 
 `mode: run` (with `api-key`, `model` and `budget`) runs the agent; it spends money and needs a secret, so the action
@@ -186,7 +201,7 @@ refuses it on pull requests: run it on `workflow_dispatch` or a schedule only.
 # .pre-commit-config.yaml (after `obey enforce --targets precommit`)
 repos:
   - repo: https://github.com/B0yko/obey
-    rev: v0.1.0
+    rev: v0.1.1
     hooks:
       - id: obey
       - id: obey-commit-msg
@@ -201,8 +216,8 @@ The badge: `obey lint` and `obey run` write `obey-badge.json` in the
 The study's agent runs were made for this release on 2026-10-23 and 2026-10-24 (UTC) on a Mac Studio M4 Max, 128 GB
 (Docker in colima), with `anthropic/claude-haiku-5.5` through OpenRouter in Claude Code 2.1.309 and OpenCode 1.19.4.
 The survey and the extraction and checker evaluations involve no agent runs.
-Full report: [study/REPORT.md](study/REPORT.md). Pre-registration: [study/PREREGISTRATION.md](study/PREREGISTRATION.md).
-Every change after it: [study/DEVIATIONS.md](study/DEVIATIONS.md).
+Full report: [study/REPORT.md](https://github.com/B0yko/obey/blob/main/study/REPORT.md). Pre-registration: [study/PREREGISTRATION.md](https://github.com/B0yko/obey/blob/main/study/PREREGISTRATION.md).
+Every change after it: [study/DEVIATIONS.md](https://github.com/B0yko/obey/blob/main/study/DEVIATIONS.md).
 
 ### What public AGENTS.md files say (600 repositories)
 
@@ -220,8 +235,8 @@ phrasings, 14 runs per cell. This is **probe compliance**: on small tasks built
 route, not compliance in everyday use.
 
 <picture>
-  <source media="(prefers-color-scheme: dark)" srcset="docs/images/compliance-by-phrasing-dark.svg">
-  <img alt="Probe compliance by phrasing and harness: about 80% with no rule, 100% with the rule in any phrasing" src="docs/images/compliance-by-phrasing-light.svg">
+  <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/compliance-by-phrasing-dark.svg">
+  <img alt="Probe compliance by phrasing and harness: about 80% with no rule, 100% with the rule in any phrasing" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/compliance-by-phrasing-light.svg">
 </picture>
 
 | Condition | As registered | Corrected | Exercised runs |
@@ -257,8 +272,8 @@ git pre-commit check, and the same probes were run again, still without the rule
 rule only when a hook refuses it.
 
 <picture>
-  <source media="(prefers-color-scheme: dark)" srcset="docs/images/enforcement-before-after-dark.svg">
-  <img alt="Probe compliance without the rule, before and with hooks, for the five rule types agents broke" src="docs/images/enforcement-before-after-light.svg">
+  <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/enforcement-before-after-dark.svg">
+  <img alt="Probe compliance without the rule, before and with hooks, for the five rule types agents broke" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/enforcement-before-after-light.svg">
 </picture>
 
 - In the 13 cells (rule type × repository × agent) that agents broke without the rule, compliance went from 34%
@@ -364,8 +379,8 @@ differ (checked on 2026-10-23):
   than from your instruction file.
 - [nunchi](https://github.com/seob717/nunchi) ([article](https://braindetox.kr/en/posts/claude_md_hooks_compile_2026.html))
   compiles CLAUDE.md rules into trigger-bound rules that a hook re-delivers when an action fires.
-- [agents-md-evals](https://github.com/vltansky/agents-md-evals) A/B tests a Claude instruction file present versus
-  removed, using assertions from the repository's history.
+- [agents-md-evals](https://github.com/vltansky/agents-md-evals) (archived) A/B tests a Claude instruction file
+  present versus removed, using assertions from the repository's history.
 - [ruleProb](https://github.com/canblmz1/ruleProb) extracts rules from CLAUDE.md, AGENTS.md and Cursor rules, builds
   sandbox scenarios, runs a provider against each and scores commands, protected writes and code patterns; it is the
   closest in purpose. obey differs in its probes-as-temptations with a no-rule baseline, per-rule rates with intervals

{obey_cli-0.1.0 → obey_cli-0.1.1}/README.md

@@ -1,37 +1,52 @@
+<p align="center">
+  <img alt="obey, unit tests for your AGENTS.md: a report card of the rules a coding agent followed, and a hook that fixed the one it broke" src="https://raw.githubusercontent.com/B0yko/obey/main/.github/social-preview.png" width="860">
+</p>
+
 # obey
 
-**Unit tests for AGENTS.md.** obey turns the rules in your agent instruction files into checks, runs your coding agent
-on small tempting tasks in a sandbox, measures which rules it actually follows from what it did, and compiles the ones it
-ignores into hooks for Claude Code, Codex CLI, Cursor, Gemini CLI, Copilot CLI and OpenCode.
+**You wrote rules for your coding agent. Does it follow them?**
+
+AGENTS.md, CLAUDE.md and Cursor rules tell coding agents how to work in your repository, but nothing checks that they
+do. obey tests every rule against what the agent actually did, and turns the rules it breaks into hooks it cannot skip.
 
 [![CI](https://github.com/B0yko/obey/actions/workflows/ci.yml/badge.svg)](https://github.com/B0yko/obey/actions/workflows/ci.yml)
 [![Release](https://img.shields.io/github/v/release/B0yko/obey)](https://github.com/B0yko/obey/releases)
-[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](LICENSE)
+[![PyPI](https://img.shields.io/pypi/v/obey-cli)](https://pypi.org/project/obey-cli/)
+[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](https://github.com/B0yko/obey/blob/main/LICENSE)
 
-**On tempting tasks, Claude Haiku 5.5 in Claude Code and OpenCode followed the rule under test in all 1,474 exercised
-runs where it was in AGENTS.md, against 81% without it (stratified difference +17.6 points [95% CI 4.9–34.1]); the
-phrasing (imperative, prohibition, or with a reason) made no measurable difference, since every phrasing was at
-100%. Hooks without the rule raised compliance from 34% to 100% for the rules agents broke, and in 0 of 123 blocked
-runs did the agent break the rule another way.** This is probe compliance, measured after correcting 136 checker
-false positives (94% before the correction); see [how it was measured](study/REPORT.md).
+1. **Check.** `obey init` reads your instruction files and turns rules such as "use pnpm", "never push" or "run the
+   tests before you finish" into automatic checks.
+2. **Test.** `obey run` gives your agent a small task per rule where breaking it is the easy way out, in a sandbox,
+   and scores what it did (commands, diffs, commits), not what it said. You get a report per rule.
+3. **Enforce.** `obey enforce` writes hooks for Claude Code, Codex CLI, OpenCode, Copilot CLI, Gemini CLI and Cursor,
+   plus a git pre-commit check, so a broken rule is blocked the moment the agent tries.
 
-<picture>
-  <img alt="obey demo replaying recorded runs: per-rule compliance and the before/after table with hooks" src="docs/images/demo.svg">
-</picture>
+**What we found** in 2,663 sandboxed runs of Claude Haiku 5.5 in Claude Code and OpenCode, on test tasks built to tempt
+the agent: it followed a rule every time the rule was in AGENTS.md (1,474 of 1,474 runs) and 81% of the time when it
+was not. How the rule was worded made no measurable difference. For the rules it broke without the rule written down,
+a hook alone raised compliance from 34% to 100%, and in 0 of 123 blocked runs did the agent find a way around it.
+[How it was measured, the caveats, and the 136 errors we found in our own checkers](#results).
+
+### Try it in one command
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey demo
+uvx --from obey-cli obey demo
 ```
 
-The demo needs no API key and no Docker: it replays agent runs recorded in the study and scores them again.
+No API key and no Docker: the demo replays agent runs recorded in the study and scores them again.
+
+<picture>
+  <img alt="obey demo replaying recorded runs: per-rule compliance and the before/after table with hooks" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/demo.svg">
+</picture>
 
 ## Quickstart
 
-In your repository (Python 3.10+, git, [uv](https://docs.astral.sh/uv/)):
+In your repository (Python 3.10+, git, [uv](https://docs.astral.sh/uv/)). The package on PyPI is `obey-cli`; to get a
+plain `obey` command, install it with `uv tool install obey-cli` or `pipx install obey-cli`.
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey init      # extract rules into obey.yaml and create probes
-uvx --from git+https://github.com/B0yko/obey obey lint      # no agent runs: checks rules, hashes, enforcement; writes the badge
+uvx --from obey-cli obey init      # extract rules into obey.yaml and create probes
+uvx --from obey-cli obey lint      # no agent runs: checks rules, hashes, enforcement; writes the badge
 ```
 
 Read `obey.yaml`: every rule has a checker, or a status that says why not (`uncheckable` for rules that need taste,
@@ -39,23 +54,23 @@ Read `obey.yaml`: every rule has a checker, or a status that says why not (`unch
 
 ```bash
 # enforce now, no runs needed: hooks for every agent plus a git pre-commit check
-uvx --from git+https://github.com/B0yko/obey obey enforce --dry-run
-uvx --from git+https://github.com/B0yko/obey obey enforce
+uvx --from obey-cli obey enforce --dry-run
+uvx --from obey-cli obey enforce
 
 # measure: needs Docker (or colima) and an OpenRouter key; the key never enters the sandbox
-uvx --from git+https://github.com/B0yko/obey obey doctor --build-image          # once: the sandbox image, a few GB
+uvx --from obey-cli obey doctor --build-image          # once: the sandbox image, a few GB
 export OPENROUTER_API_KEY=...
-uvx --from git+https://github.com/B0yko/obey obey run --estimate --model anthropic/claude-haiku-5.5
-uvx --from git+https://github.com/B0yko/obey obey run --model anthropic/claude-haiku-5.5 --budget 2
-uvx --from git+https://github.com/B0yko/obey obey enforce --from .obey/report.json   # hooks for the rules it broke
+uvx --from obey-cli obey run --estimate --model anthropic/claude-haiku-5.5
+uvx --from obey-cli obey run --model anthropic/claude-haiku-5.5 --budget 2
+uvx --from obey-cli obey enforce --from .obey/report.json   # hooks for the rules it broke
 cp .obey/report.json before.json
-uvx --from git+https://github.com/B0yko/obey obey run --compare before.json --model anthropic/claude-haiku-5.5
+uvx --from obey-cli obey run --compare before.json --model anthropic/claude-haiku-5.5
 ```
 
 ### Test your AGENTS.md in 10 minutes
 
 ```bash
-uvx --from git+https://github.com/B0yko/obey obey run --model anthropic/claude-haiku-5.5 --trials 3 --baseline-trials 1 --budget 2
+uvx --from obey-cli obey run --model anthropic/claude-haiku-5.5 --trials 3 --baseline-trials 1 --budget 2
 ```
 
 On the demo repository (11 checkable rules with probes) this made 44 runs of Claude Code 2.1.309 in 9 min 20 s and
@@ -120,12 +135,12 @@ from the vendor's documentation. Every target also gets the git pre-commit
 check. Codex CLI loads project hooks only after you trust them (`/hooks`); Copilot CLI's prompt mode needs
 `--trust-folder` or `GITHUB_COPILOT_PROMPT_MODE_REPO_HOOKS=true`.
 
-Details, flags and isolation checks: [docs/agents.md](docs/agents.md). Checker definitions:
-[docs/checkers.md](docs/checkers.md). Enforcement targets: [docs/enforce.md](docs/enforce.md).
+Details, flags and isolation checks: [docs/agents.md](https://github.com/B0yko/obey/blob/main/docs/agents.md). Checker definitions:
+[docs/checkers.md](https://github.com/B0yko/obey/blob/main/docs/checkers.md). Enforcement targets: [docs/enforce.md](https://github.com/B0yko/obey/blob/main/docs/enforce.md).
 
 ## Configuration
 
-`obey.yaml` is the reviewed list of rules ([reference](docs/config.md)). Environment variables:
+`obey.yaml` is the reviewed list of rules ([reference](https://github.com/B0yko/obey/blob/main/docs/config.md)). Environment variables:
 
 | Variable | Used by | Meaning |
 |---|---|---|
@@ -145,7 +160,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
-      - uses: B0yko/obey@v0.1.0
+      - uses: B0yko/obey@v0.1.1
 ```
 
 `mode: run` (with `api-key`, `model` and `budget`) runs the agent; it spends money and needs a secret, so the action
@@ -155,7 +170,7 @@ refuses it on pull requests: run it on `workflow_dispatch` or a schedule only.
 # .pre-commit-config.yaml (after `obey enforce --targets precommit`)
 repos:
   - repo: https://github.com/B0yko/obey
-    rev: v0.1.0
+    rev: v0.1.1
     hooks:
       - id: obey
       - id: obey-commit-msg
@@ -170,8 +185,8 @@ The badge: `obey lint` and `obey run` write `obey-badge.json` in the
 The study's agent runs were made for this release on 2026-10-23 and 2026-10-24 (UTC) on a Mac Studio M4 Max, 128 GB
 (Docker in colima), with `anthropic/claude-haiku-5.5` through OpenRouter in Claude Code 2.1.309 and OpenCode 1.19.4.
 The survey and the extraction and checker evaluations involve no agent runs.
-Full report: [study/REPORT.md](study/REPORT.md). Pre-registration: [study/PREREGISTRATION.md](study/PREREGISTRATION.md).
-Every change after it: [study/DEVIATIONS.md](study/DEVIATIONS.md).
+Full report: [study/REPORT.md](https://github.com/B0yko/obey/blob/main/study/REPORT.md). Pre-registration: [study/PREREGISTRATION.md](https://github.com/B0yko/obey/blob/main/study/PREREGISTRATION.md).
+Every change after it: [study/DEVIATIONS.md](https://github.com/B0yko/obey/blob/main/study/DEVIATIONS.md).
 
 ### What public AGENTS.md files say (600 repositories)
 
@@ -189,8 +204,8 @@ phrasings, 14 runs per cell. This is **probe compliance**: on small tasks built
 route, not compliance in everyday use.
 
 <picture>
-  <source media="(prefers-color-scheme: dark)" srcset="docs/images/compliance-by-phrasing-dark.svg">
-  <img alt="Probe compliance by phrasing and harness: about 80% with no rule, 100% with the rule in any phrasing" src="docs/images/compliance-by-phrasing-light.svg">
+  <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/compliance-by-phrasing-dark.svg">
+  <img alt="Probe compliance by phrasing and harness: about 80% with no rule, 100% with the rule in any phrasing" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/compliance-by-phrasing-light.svg">
 </picture>
 
 | Condition | As registered | Corrected | Exercised runs |
@@ -226,8 +241,8 @@ git pre-commit check, and the same probes were run again, still without the rule
 rule only when a hook refuses it.
 
 <picture>
-  <source media="(prefers-color-scheme: dark)" srcset="docs/images/enforcement-before-after-dark.svg">
-  <img alt="Probe compliance without the rule, before and with hooks, for the five rule types agents broke" src="docs/images/enforcement-before-after-light.svg">
+  <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/enforcement-before-after-dark.svg">
+  <img alt="Probe compliance without the rule, before and with hooks, for the five rule types agents broke" src="https://raw.githubusercontent.com/B0yko/obey/main/docs/images/enforcement-before-after-light.svg">
 </picture>
 
 - In the 13 cells (rule type × repository × agent) that agents broke without the rule, compliance went from 34%
@@ -333,8 +348,8 @@ differ (checked on 2026-10-23):
   than from your instruction file.
 - [nunchi](https://github.com/seob717/nunchi) ([article](https://braindetox.kr/en/posts/claude_md_hooks_compile_2026.html))
   compiles CLAUDE.md rules into trigger-bound rules that a hook re-delivers when an action fires.
-- [agents-md-evals](https://github.com/vltansky/agents-md-evals) A/B tests a Claude instruction file present versus
-  removed, using assertions from the repository's history.
+- [agents-md-evals](https://github.com/vltansky/agents-md-evals) (archived) A/B tests a Claude instruction file
+  present versus removed, using assertions from the repository's history.
 - [ruleProb](https://github.com/canblmz1/ruleProb) extracts rules from CLAUDE.md, AGENTS.md and Cursor rules, builds
   sandbox scenarios, runs a provider against each and scores commands, protected writes and code patterns; it is the
   closest in purpose. obey differs in its probes-as-temptations with a no-rule baseline, per-rule rates with intervals

{obey_cli-0.1.0 → obey_cli-0.1.1}/action.yml

@@ -13,7 +13,7 @@ inputs:
     default: "."
   obey-version:
     description: obey version to install (a git tag of github.com/B0yko/obey)
-    default: v0.1.0
+    default: v0.1.1
   agent:
     description: "run mode: agent to run (claude-code, codex, opencode, copilot, gemini)"
     default: claude-code

{obey_cli-0.1.0 → obey_cli-0.1.1}/docs/config.md

@@ -5,7 +5,7 @@
 
 ```yaml
 version: 1
-generated_by: obey 0.1.0
+generated_by: obey 0.1.1
 sources:                       # instruction files at extraction time; obey lint re-checks their hashes
   - file: AGENTS.md
     sha256: 3b1f…