rsi-loop 0.1.0__tar.gz

rsi_loop-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,29 @@
+name: Publish to PyPI
+
+on:
+  release:
+    types: [published]
+
+permissions:
+  id-token: write  # Required for trusted publishing
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    environment: pypi
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install build tools
+        run: pip install build
+
+      - name: Build package
+        run: python -m build
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

rsi_loop-0.1.0/.gitignore

@@ -0,0 +1,7 @@
+__pycache__/
+*.pyc
+.coverage
+*.egg-info/
+dist/
+build/
+.venv/

rsi_loop-0.1.0/CHANGELOG.md

@@ -0,0 +1,19 @@
+# Changelog
+
+## [0.1.0] - 2026-02-22
+
+### Added
+- Initial release
+- Core modules: Observer, Analyzer, Fixer, RSILoop
+- Type definitions: Outcome, Pattern, Fix, Config
+- Auto-classification of error messages
+- Recurrence detection (3+ occurrences in window)
+- Health scoring (0.0–1.0, recency-weighted)
+- Cross-source correlation for related issues
+- Error message clustering with normalization
+- Safe auto-fix for configurable categories
+- Fix proposal generation for unsafe categories
+- Background loop with configurable interval
+- Integrations: Claude Code adapter, generic file-based adapter, webhook adapter
+- Full test suite with 85%+ coverage
+- CI with Python 3.10/3.11/3.12, pytest, ruff

rsi_loop-0.1.0/LICENSE

@@ -0,0 +1,189 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work.
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to the Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by the Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding any notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   Copyright 2026 ClawInfra
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

rsi_loop-0.1.0/PKG-INFO

@@ -0,0 +1,151 @@
+Metadata-Version: 2.4
+Name: rsi-loop
+Version: 0.1.0
+Summary: Universal self-improvement loop for AI agents. Observe → Analyze → Fix → Verify.
+Project-URL: Homepage, https://github.com/clawinfra/rsi-loop
+Project-URL: Repository, https://github.com/clawinfra/rsi-loop
+Project-URL: Issues, https://github.com/clawinfra/rsi-loop/issues
+Author-email: ClawInfra <alex.chen31337@gmail.com>
+License-Expression: Apache-2.0
+License-File: LICENSE
+Keywords: agents,ai,feedback-loop,rsi,self-improvement
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Provides-Extra: dev
+Requires-Dist: pytest-cov>=5.0; extra == 'dev'
+Requires-Dist: pytest>=8.0; extra == 'dev'
+Requires-Dist: ruff>=0.4; extra == 'dev'
+Provides-Extra: webhook
+Requires-Dist: flask>=3.0; extra == 'webhook'
+Description-Content-Type: text/markdown
+
+# 🔄 RSI Loop
+
+[![CI](https://github.com/clawinfra/rsi-loop/actions/workflows/ci.yml/badge.svg)](https://github.com/clawinfra/rsi-loop/actions/workflows/ci.yml)
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
+[![Python](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://python.org)
+
+**Every AI agent makes mistakes. RSI Loop makes them learn.**
+
+---
+
+## The Problem
+
+AI agents repeat the same failures. They hit rate limits, return empty responses, lose context, pick the wrong model — and do it all again next session. There's no feedback loop. No memory of what went wrong. No automatic improvement.
+
+RSI Loop is the missing primitive: a universal **recursive self-improvement** loop that works with *any* agent framework.
+
+## How It Works
+
+```
+    ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐
+    │ OBSERVE  │────▶│ ANALYZE  │────▶│   FIX    │────▶│  VERIFY  │
+    │          │     │          │     │          │     │          │
+    │ Record   │     │ Detect   │     │ Generate │     │ Check    │
+    │ outcomes │     │ patterns │     │ & apply  │     │ health   │
+    └──────────┘     └──────────┘     └──────────┘     └────┬─────┘
+         ▲                                                   │
+         └───────────────────────────────────────────────────┘
+```
+
+1. **Observe** — Record task outcomes: success/failure, quality, errors, model used, duration
+2. **Analyze** — Detect patterns: recurring failures, error clusters, cross-source correlations
+3. **Fix** — Auto-fix safe categories (routing, retries, thresholds); propose fixes for the rest
+4. **Verify** — Track health score over time; confirm fixes reduced failure rates
+
+## Quick Start
+
+```bash
+uv add rsi-loop
+```
+
+> **Not using uv?** `uv add rsi-loop` works too.
+
+```python
+from rsi_loop import RSILoop
+
+loop = RSILoop()
+loop.observer.record_simple("code_generation", success=True, model="sonnet-4.6")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+
+# Run improvement cycle
+patterns = loop.run_cycle()
+print(f"Health: {loop.health_score():.0%}")
+print(f"Patterns found: {len(patterns)}")
+```
+
+## Framework Integrations
+
+### Claude Code / OpenClaw
+
+```python
+from rsi_loop.integrations.claude_code import ClaudeCodeAdapter
+
+adapter = ClaudeCodeAdapter(data_dir="./rsi_data")
+adapter.on_task_complete("code_review", success=True, model="claude-sonnet-4")
+adapter.on_model_fallback("claude-opus-4", "claude-sonnet-4", "rate_limit")
+adapter.on_session_reset()
+```
+
+### Generic (File-Based)
+
+Drop JSON files into a watch directory — works with any agent:
+
+```python
+from rsi_loop.integrations.generic import GenericAdapter
+
+adapter = GenericAdapter(watch_dir="./rsi_inbox")
+# Your agent writes: {"task": "search", "success": false, "error": "timeout"}
+# RSI picks it up automatically
+outcomes = adapter.poll()
+```
+
+### Webhook
+
+```python
+from rsi_loop.integrations.webhook import WebhookAdapter
+
+app = WebhookAdapter(data_dir="./rsi_data").create_app()
+# POST /observe {"task": "code_gen", "success": true, "quality": 4}
+# GET /health
+# GET /patterns
+```
+
+## Features
+
+- **Auto-classification** — Categorizes errors automatically: rate_limit, empty_response, timeout, context_loss, etc.
+- **Recurrence detection** — Flags issues that repeat 3+ times within the analysis window
+- **Health scoring** — 0.0 (broken) to 1.0 (healthy), recency-weighted
+- **Cross-source correlation** — Detects related issues across different agent sources
+- **Error clustering** — Groups similar error messages even with different IDs/numbers
+- **Safe auto-fix** — Automatically applies fixes for safe categories (routing, retries, thresholds)
+- **Fix proposals** — Generates detailed proposals for unsafe categories, saved for human review
+- **Background loop** — Run continuous improvement cycles in a background thread
+- **Framework-agnostic** — Works with Claude Code, Cursor, Codex, or any custom agent
+- **Zero dependencies** — Core package has no external dependencies (integrations optional)
+
+## Documentation
+
+- [Architecture](docs/architecture.md) — How RSI Loop works internally
+- [Source Taxonomy](docs/sources.md) — Source classification and cross-source correlation
+- [Auto-Fix](docs/auto_fix.md) — How auto-fix works, safe categories, and fix templates
+
+## Examples
+
+- [Claude Code Setup](examples/claude_code_setup.md)
+- [Cursor Setup](examples/cursor_setup.md)
+- [Generic Agent Setup](examples/generic_agent_setup.md)
+- [Quick Start Script](examples/quick_start.py)
+
+---
+
+Built by [ClawInfra](https://github.com/clawinfra) — infrastructure for autonomous agents.

rsi_loop-0.1.0/README.md

@@ -0,0 +1,123 @@
+# 🔄 RSI Loop
+
+[![CI](https://github.com/clawinfra/rsi-loop/actions/workflows/ci.yml/badge.svg)](https://github.com/clawinfra/rsi-loop/actions/workflows/ci.yml)
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
+[![Python](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://python.org)
+
+**Every AI agent makes mistakes. RSI Loop makes them learn.**
+
+---
+
+## The Problem
+
+AI agents repeat the same failures. They hit rate limits, return empty responses, lose context, pick the wrong model — and do it all again next session. There's no feedback loop. No memory of what went wrong. No automatic improvement.
+
+RSI Loop is the missing primitive: a universal **recursive self-improvement** loop that works with *any* agent framework.
+
+## How It Works
+
+```
+    ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐
+    │ OBSERVE  │────▶│ ANALYZE  │────▶│   FIX    │────▶│  VERIFY  │
+    │          │     │          │     │          │     │          │
+    │ Record   │     │ Detect   │     │ Generate │     │ Check    │
+    │ outcomes │     │ patterns │     │ & apply  │     │ health   │
+    └──────────┘     └──────────┘     └──────────┘     └────┬─────┘
+         ▲                                                   │
+         └───────────────────────────────────────────────────┘
+```
+
+1. **Observe** — Record task outcomes: success/failure, quality, errors, model used, duration
+2. **Analyze** — Detect patterns: recurring failures, error clusters, cross-source correlations
+3. **Fix** — Auto-fix safe categories (routing, retries, thresholds); propose fixes for the rest
+4. **Verify** — Track health score over time; confirm fixes reduced failure rates
+
+## Quick Start
+
+```bash
+uv add rsi-loop
+```
+
+> **Not using uv?** `uv add rsi-loop` works too.
+
+```python
+from rsi_loop import RSILoop
+
+loop = RSILoop()
+loop.observer.record_simple("code_generation", success=True, model="sonnet-4.6")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+loop.observer.record_simple("api_call", success=False, error="429 Too Many Requests")
+
+# Run improvement cycle
+patterns = loop.run_cycle()
+print(f"Health: {loop.health_score():.0%}")
+print(f"Patterns found: {len(patterns)}")
+```
+
+## Framework Integrations
+
+### Claude Code / OpenClaw
+
+```python
+from rsi_loop.integrations.claude_code import ClaudeCodeAdapter
+
+adapter = ClaudeCodeAdapter(data_dir="./rsi_data")
+adapter.on_task_complete("code_review", success=True, model="claude-sonnet-4")
+adapter.on_model_fallback("claude-opus-4", "claude-sonnet-4", "rate_limit")
+adapter.on_session_reset()
+```
+
+### Generic (File-Based)
+
+Drop JSON files into a watch directory — works with any agent:
+
+```python
+from rsi_loop.integrations.generic import GenericAdapter
+
+adapter = GenericAdapter(watch_dir="./rsi_inbox")
+# Your agent writes: {"task": "search", "success": false, "error": "timeout"}
+# RSI picks it up automatically
+outcomes = adapter.poll()
+```
+
+### Webhook
+
+```python
+from rsi_loop.integrations.webhook import WebhookAdapter
+
+app = WebhookAdapter(data_dir="./rsi_data").create_app()
+# POST /observe {"task": "code_gen", "success": true, "quality": 4}
+# GET /health
+# GET /patterns
+```
+
+## Features
+
+- **Auto-classification** — Categorizes errors automatically: rate_limit, empty_response, timeout, context_loss, etc.
+- **Recurrence detection** — Flags issues that repeat 3+ times within the analysis window
+- **Health scoring** — 0.0 (broken) to 1.0 (healthy), recency-weighted
+- **Cross-source correlation** — Detects related issues across different agent sources
+- **Error clustering** — Groups similar error messages even with different IDs/numbers
+- **Safe auto-fix** — Automatically applies fixes for safe categories (routing, retries, thresholds)
+- **Fix proposals** — Generates detailed proposals for unsafe categories, saved for human review
+- **Background loop** — Run continuous improvement cycles in a background thread
+- **Framework-agnostic** — Works with Claude Code, Cursor, Codex, or any custom agent
+- **Zero dependencies** — Core package has no external dependencies (integrations optional)
+
+## Documentation
+
+- [Architecture](docs/architecture.md) — How RSI Loop works internally
+- [Source Taxonomy](docs/sources.md) — Source classification and cross-source correlation
+- [Auto-Fix](docs/auto_fix.md) — How auto-fix works, safe categories, and fix templates
+
+## Examples
+
+- [Claude Code Setup](examples/claude_code_setup.md)
+- [Cursor Setup](examples/cursor_setup.md)
+- [Generic Agent Setup](examples/generic_agent_setup.md)
+- [Quick Start Script](examples/quick_start.py)
+
+---
+
+Built by [ClawInfra](https://github.com/clawinfra) — infrastructure for autonomous agents.

rsi_loop-0.1.0/docs/architecture.md

@@ -0,0 +1,72 @@
+# Architecture
+
+RSI Loop implements a four-phase recursive self-improvement cycle.
+
+## Components
+
+### Observer
+Records task outcomes to a JSONL store. Each outcome captures: task type, success/failure, quality (1-5), error messages, model used, duration, and arbitrary tags/metadata.
+
+**Auto-classification**: Error messages are automatically classified into issue types (rate_limit, timeout, empty_response, context_loss, etc.) using keyword matching.
+
+**Storage**: Outcomes are appended to `{data_dir}/outcomes.jsonl` — one JSON object per line.
+
+### Analyzer
+Scans recorded outcomes and detects improvement patterns.
+
+**Pattern detection** works two ways:
+1. **Task/issue grouping** — Groups outcomes by `(task_type, issue)` pairs. Patterns with ≥2 occurrences (or ≥1 for high-severity issues) are surfaced.
+2. **Error clustering** — Normalizes error messages (strips IDs, numbers) and groups similar errors. Catches patterns that span multiple task types.
+
+**Health score**: `success_rate × (avg_quality / 5.0)`, computed over the analysis window. Range: 0.0 (broken) to 1.0 (healthy).
+
+**Cross-source correlation**: Detects when related issues (e.g., session_reset + context_loss) appear across different agent sources, suggesting systemic problems.
+
+**Recurrence detection**: Compares current patterns against the previous analysis cycle. Recurring patterns are flagged with trend direction (increasing/stable).
+
+### Fixer
+Generates fix proposals for detected patterns.
+
+**Safe categories** (configurable, default: routing_config, threshold_tuning, retry_logic) get auto-applied status. Everything else produces a draft proposal saved to `{data_dir}/proposals/` for human review.
+
+### RSILoop
+The main orchestrator. Ties Observer, Analyzer, and Fixer together. Provides:
+- `run_cycle()` — One full observe → analyze → fix pass
+- `health_score()` — Current health
+- `start_background(interval)` — Background improvement loop in a daemon thread
+
+## Data Flow
+
+```
+Agent task completes
+    │
+    ▼
+Observer.record_simple(task, success, error, model)
+    │ auto-classify issues from error message
+    │ append to outcomes.jsonl
+    ▼
+Analyzer.analyze()
+    │ load outcomes from window
+    │ group by (task_type, issue)
+    │ cluster similar errors
+    │ detect recurrences
+    │ compute health score
+    ▼
+Fixer.propose_and_apply(pattern)
+    │ generate fix proposal
+    │ auto-apply if safe category
+    │ save proposal to proposals/
+    ▼
+Health score improves (or doesn't → next cycle catches it)
+```
+
+## Storage Layout
+
+```
+{data_dir}/
+├── outcomes.jsonl    # Raw outcome records (append-only)
+├── patterns.json     # Latest analysis results (overwritten each cycle)
+└── proposals/        # Fix proposals (one JSON file per fix)
+    ├── abc12345.json
+    └── def67890.json
+```

rsi_loop-0.1.0/docs/auto_fix.md

@@ -0,0 +1,78 @@
+# Auto-Fix
+
+RSI Loop can automatically apply fixes for "safe" categories — changes that are low-risk and well-understood.
+
+## Safe Categories
+
+By default, three fix types are considered safe for auto-application:
+
+| Category | Description | Example |
+|----------|-------------|---------|
+| `routing_config` | Model routing changes | Adjusting fallback chains |
+| `threshold_tuning` | Timeout/threshold adjustments | Increasing request timeouts |
+| `retry_logic` | Adding/adjusting retry behavior | Backoff for rate-limited endpoints |
+
+Configure via `Config.safe_categories`:
+
+```python
+from rsi_loop import RSILoop, Config
+
+loop = RSILoop(Config(
+    safe_categories=["routing_config", "retry_logic", "threshold_tuning", "investigation"]
+))
+```
+
+## Fix Templates
+
+Each issue type maps to a fix template:
+
+| Issue | Fix Type | Action |
+|-------|----------|--------|
+| rate_limit | retry_logic | Add/increase retry backoff |
+| model_fallback | routing_config | Update fallback chain |
+| wrong_model_tier | routing_config | Adjust tier thresholds |
+| cost_overrun | routing_config | Lower cost ceiling |
+| slow_response | threshold_tuning | Increase timeouts |
+| timeout | threshold_tuning | Increase timeouts |
+| empty_response | retry_logic | Add empty-response detection and retry |
+| session_reset | investigation | Investigate (not auto-fixable by default) |
+
+## How It Works
+
+1. `Fixer.propose(pattern)` generates a `Fix` with changes and a `safe_category`
+2. `Fixer.apply_if_safe(fix)` checks if `safe_category` is in `Config.safe_categories`
+3. If safe → status set to "applied", proposal saved
+4. If not safe → status stays "draft", proposal saved for human review
+
+**Important**: Even "applied" fixes only update their status in the proposal file. Actual code/config changes require agent-specific implementation — RSI Loop tells you *what* to fix, your agent decides *how*.
+
+## Proposals
+
+All fix proposals are saved as JSON files in `{data_dir}/proposals/`:
+
+```json
+{
+  "id": "abc12345",
+  "pattern_id": "api_call-rate_lim-5",
+  "type": "auto",
+  "status": "applied",
+  "target": "relevant integration code",
+  "changes": [
+    {
+      "target": "relevant integration code",
+      "action": "Add retry with backoff for 'rate_limit' errors",
+      "detail": "Detected 5x occurrences"
+    }
+  ],
+  "safe_category": "retry_logic",
+  "description": "Add/increase retry backoff for rate-limited endpoints"
+}
+```
+
+## Disabling Auto-Fix
+
+```python
+loop = RSILoop(Config(auto_fix_enabled=False))
+# Or set safe_categories to empty:
+loop = RSILoop(Config(safe_categories=[]))
+```