Skip to content

The agent that turns websites into APIs!

fix: address pr comments

kalil03217ffe8331c2e8
diff --git a/CLAUDE.md b/CLAUDE.md
index 0468589..82f2f6b 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -191,9 +191,9 @@ src/reverse_api/
├── config.json # User configuration
├── history.json # Run history with costs
└── runs/
- ├── har/{id} # Captured traffic
- ├── scripts/{id} # Generated API clients
- └── messages/{id}.jsonl # LLM conversation logs
+ ├── har/{run_id} # Captured traffic
+ ├── scripts/{run_id} # Generated API clients
+ └── messages/{run_id}.jsonl # LLM conversation logs
```
## Dependencies Management
diff --git a/plugins/reverse-api-engineer/CHANGELOG.md b/plugins/reverse-api-engineer/CHANGELOG.md
index cb1f972..2a2774c 100644
--- a/plugins/reverse-api-engineer/CHANGELOG.md
+++ b/plugins/reverse-api-engineer/CHANGELOG.md
@@ -5,7 +5,7 @@ All notable changes to the Reverse API Engineer plugin will be documented in thi
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
-## [1.0.0] - 2026-01-01
+## [1.0.0] - 2026-01-02
### Added
- Initial release of Reverse API Engineer plugin
@@ -21,11 +21,3 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- Integration with rae-playwright-mcp for browser control
- UUID generation and run history tracking
- Support for both run_id and direct HAR path references
-
-### Features
-- Generate production-ready Python API clients from browser traffic
-- Automatic filtering of irrelevant HAR entries (assets, analytics, ads)
-- Support for multiple authentication patterns (Bearer, API keys, OAuth)
-- Error handling and logging in generated code
-- Type hints and comprehensive documentation
-- Session management for connection reuse
diff --git a/plugins/reverse-api-engineer/README.md b/plugins/reverse-api-engineer/README.md
index 846c850..eb7fb34 100644
--- a/plugins/reverse-api-engineer/README.md
+++ b/plugins/reverse-api-engineer/README.md
@@ -58,7 +58,7 @@ Re-run AI generation on a previous HAR capture.
**Examples:**
```
/reverse-api-engineer:engineer abc-123-def
-/reverse-api-engineer:engineer ~/.reverse-api/runs/xyz-789/har/network.har
+/reverse-api-engineer:engineer ~/.reverse-api/runs/har/xyz-789/recording.har
/reverse-api-engineer:engineer ./captures/my-traffic.har
```
@@ -115,7 +115,7 @@ This skill automatically activates when you mention API reverse engineering task
Captured traffic is saved to:
```
-~/.reverse-api/runs/har/{run_id}/network.har
+~/.reverse-api/runs/har/{run_id}/recording.har
```
### Generated Scripts
@@ -186,7 +186,7 @@ Scripts saved to: ./scripts/apple_jobs_api/
### Example 2: Re-engineer Existing Capture
```
-You: /reverse-api-engineer:engineer ~/.reverse-api/runs/abc-123/har/network.har
+You: /reverse-api-engineer:engineer ~/.reverse-api/runs/abc-123/har/recording.har
[AI analyzes HAR]
Found endpoints:
diff --git a/plugins/reverse-api-engineer/agents/api-reverse-engineer.md b/plugins/reverse-api-engineer/agents/api-reverse-engineer.md
index 1202a23..c5143fa 100644
--- a/plugins/reverse-api-engineer/agents/api-reverse-engineer.md
+++ b/plugins/reverse-api-engineer/agents/api-reverse-engineer.md
@@ -134,7 +134,7 @@ Since you have a HAR file, use engineer mode to analyze it:
/reverse-api-engineer:engineer <path-to-har-file>
-Example: /reverse-api-engineer:engineer ~/.reverse-api/runs/abc123/har/network.har
+Example: /reverse-api-engineer:engineer ~/.reverse-api/runs/abc123/har/recording.har
The AI will analyze the traffic and generate a production-ready Python API client.
```
diff --git a/plugins/reverse-api-engineer/commands/agent.md b/plugins/reverse-api-engineer/commands/agent.md
index 0ded431..7b49746 100644
--- a/plugins/reverse-api-engineer/commands/agent.md
+++ b/plugins/reverse-api-engineer/commands/agent.md
@@ -43,7 +43,7 @@ run_id = uuid4() (e.g., "abc-123-def-456")
Setup paths:
```
har_dir = ~/.reverse-api/runs/har/{run_id}
-har_path = {har_dir}/network.har
+har_path = {har_dir}/recording.har
```
Ensure directory exists:
@@ -61,7 +61,7 @@ Parameters:
- url: {url} or "about:blank" if no URL provided
- options:
- record_har: true
- - har_path: ~/.reverse-api/runs/har/{run_id}/network.har
+ - har_path: ~/.reverse-api/runs/har/{run_id}/recording.har
```
Inform the user:
@@ -71,7 +71,7 @@ Task: {task}
Starting URL: {url}
I'll navigate autonomously and capture all API traffic.
-HAR will be saved to: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR will be saved to: ~/.reverse-api/runs/har/{run_id}/recording.har
```
### Step 4: Autonomous Navigation
@@ -122,8 +122,8 @@ Progress update:
When navigation is complete, verify HAR file exists and contains data:
```bash
-if [ -f ~/.reverse-api/runs/har/{run_id}/network.har ]; then
- size=$(ls -lh ~/.reverse-api/runs/har/{run_id}/network.har | awk '{print $5}')
+if [ -f ~/.reverse-api/runs/har/{run_id}/recording.har ]; then
+ size=$(ls -lh ~/.reverse-api/runs/har/{run_id}/recording.har | awk '{print $5}')
echo "HAR file captured: $size"
else
echo "Error: HAR file not found"
@@ -134,7 +134,7 @@ fi
Read HAR file to count entries:
```python
import json
-with open("~/.reverse-api/runs/har/{run_id}/network.har", "r") as f:
+with open("~/.reverse-api/runs/har/{run_id}/recording.har", "r") as f:
har_data = json.load(f)
entry_count = len(har_data["log"]["entries"])
```
@@ -195,7 +195,7 @@ Generated from autonomous agent capture on {date}
Task: {task}
Run ID: {run_id}
-HAR file: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR file: ~/.reverse-api/runs/har/{run_id}/recording.har
"""
import requests
@@ -277,7 +277,7 @@ Auto-generated Python API client from autonomous browser agent.
## Generated From
- **Run ID**: {run_id}
-- **HAR File**: ~/.reverse-api/runs/har/{run_id}/network.har
+- **HAR File**: ~/.reverse-api/runs/har/{run_id}/recording.har
- **Date**: {date}
- **Task**: {task}
- **Base URL**: {base_url}
@@ -334,7 +334,7 @@ Update `~/.reverse-api/history.json` with run metadata:
"task": "{task}",
"url": "{url}",
"timestamp": "{ISO timestamp}",
- "har_path": "~/.reverse-api/runs/har/{run_id}/network.har",
+ "har_path": "~/.reverse-api/runs/har/{run_id}/recording.har",
"output_dir": "./scripts/{task_name}/",
"navigation_steps": {count},
"status": "completed"
@@ -349,7 +349,7 @@ Agent completed task successfully!
Task: {task}
Run ID: {run_id}
-HAR file: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR file: ~/.reverse-api/runs/har/{run_id}/recording.har
Agent navigation:
- Pages visited: {page_count}
diff --git a/plugins/reverse-api-engineer/commands/engineer.md b/plugins/reverse-api-engineer/commands/engineer.md
index daa2adc..dcab1a7 100644
--- a/plugins/reverse-api-engineer/commands/engineer.md
+++ b/plugins/reverse-api-engineer/commands/engineer.md
@@ -37,7 +37,7 @@ elif '/' in arg or '\\' in arg:
elif arg:
# Looks like a run_id
run_id = arg
- har_path = f"~/.reverse-api/runs/har/{run_id}/network.har"
+ har_path = f"~/.reverse-api/runs/har/{run_id}/recording.har"
else:
# No argument provided
Error: "Please provide either a run_id or path to HAR file"
@@ -61,7 +61,7 @@ If using run_id but HAR file doesn't exist:
```
Error: "No HAR file found for run_id: {run_id}
-Expected location: ~/.reverse-api/runs/har/{run_id}/network.har
+Expected location: ~/.reverse-api/runs/har/{run_id}/recording.har
Available runs:
{list runs in ~/.reverse-api/runs/}"
@@ -379,7 +379,7 @@ To regenerate with different focus:
/reverse-api-engineer:engineer abc-123-def
# Using direct path
-/reverse-api-engineer:engineer ~/.reverse-api/runs/xyz-789/har/network.har
+/reverse-api-engineer:engineer ~/.reverse-api/runs/har/xyz-789/recording.har
# Using relative path
/reverse-api-engineer:engineer ./captures/myapp.har
diff --git a/plugins/reverse-api-engineer/commands/manual.md b/plugins/reverse-api-engineer/commands/manual.md
index 30eccea..d35f6df 100644
--- a/plugins/reverse-api-engineer/commands/manual.md
+++ b/plugins/reverse-api-engineer/commands/manual.md
@@ -43,7 +43,7 @@ run_id = uuid4() (e.g., "abc-123-def-456")
Setup paths:
```
har_dir = ~/.reverse-api/runs/har/{run_id}/
-har_path = {har_dir}/network.har
+har_path = {har_dir}/recording.har
```
Ensure directory exists:
@@ -61,7 +61,7 @@ Parameters:
- url: {url} or "about:blank" if no URL provided
- options:
- record_har: true
- - har_path: ~/.reverse-api/runs/har/{run_id}/network.har
+ - har_path: ~/.reverse-api/runs/har/{run_id}/recording.har
```
Inform the user:
@@ -70,7 +70,7 @@ Browser launched with HAR recording enabled.
Navigate to the website and interact with it to trigger API calls.
Close the browser when you're done to proceed with analysis.
-HAR will be saved to: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR will be saved to: ~/.reverse-api/runs/har/{run_id}/recording.har
```
### Step 4: Wait for Browser to Close
@@ -79,7 +79,7 @@ Monitor the browser session. When it closes, the HAR file will be saved automati
Verify HAR file exists:
```bash
-ls -lh ~/.reverse-api/runs/har/{run_id}/network.har
+ls -lh ~/.reverse-api/runs/har/{run_id}/recording.har
```
If file doesn't exist or is empty:
@@ -93,7 +93,7 @@ Exit with error
Read and parse the HAR file:
```python
import json
-with open("~/.reverse-api/runs/har/{run_id}/network.har", "r") as f:
+with open("~/.reverse-api/runs/har/{run_id}/recording.har", "r") as f:
har_data = json.load(f)
```
@@ -141,7 +141,7 @@ Auto-generated API client for {domain}
Generated from HAR capture on {date}
Run ID: {run_id}
-HAR file: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR file: ~/.reverse-api/runs/har/{run_id}/recording.har
"""
import requests
@@ -223,7 +223,7 @@ Auto-generated Python API client from browser traffic capture.
## Generated From
- **Run ID**: {run_id}
-- **HAR File**: ~/.reverse-api/runs/har/{run_id}/network.har
+- **HAR File**: ~/.reverse-api/runs/har/{run_id}/recording.har
- **Date**: {date}
- **Base URL**: {base_url}
@@ -272,7 +272,7 @@ Update `~/.reverse-api/history.json` with run metadata:
"task": "{task}",
"url": "{url}",
"timestamp": "{ISO timestamp}",
- "har_path": "~/.reverse-api/runs/har/{run_id}/network.har",
+ "har_path": "~/.reverse-api/runs/har/{run_id}/recording.har",
"output_dir": "./scripts/{task_name}/",
"status": "completed"
}
@@ -285,7 +285,7 @@ Present final summary to user:
API client generated successfully!
Run ID: {run_id}
-HAR file: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR file: ~/.reverse-api/runs/har/{run_id}/recording.har
Generated files:
- ./scripts/{task_name}/api_client.py
diff --git a/plugins/reverse-api-engineer/skills/reverse-engineering-api/SKILL.md b/plugins/reverse-api-engineer/skills/reverse-engineering-api/SKILL.md
index 0bd4734..a8c79ec 100644
--- a/plugins/reverse-api-engineer/skills/reverse-engineering-api/SKILL.md
+++ b/plugins/reverse-api-engineer/skills/reverse-engineering-api/SKILL.md
@@ -34,8 +34,8 @@ This skill enables you to reverse engineer web APIs by:
When starting a browser session for API capture:
1. Launch browser with HAR recording enabled via Playwright MCP
-2. Generate a unique run ID: `{uuid}`
-3. Configure HAR output path: `~/.reverse-api/runs/har/{id}/network.har`
+2. Generate a unique run ID: `{run_id}`
+3. Configure HAR output path: `~/.reverse-api/runs/har/{run_id}/recording.har`
### During Capture
@@ -49,7 +49,7 @@ Navigate autonomously to trigger the API calls needed:
When the browser closes, note the HAR file location:
```
-HAR file saved to: ~/.reverse-api/runs/har/{run_id}/network.har
+HAR file saved to: ~/.reverse-api/runs/har/{run_id}/recording.har
```
## Phase 2: HAR Analysis
@@ -277,7 +277,7 @@ User: "Create an API client for the Apple Jobs website"
Navigate to jobs.apple.com
Perform search, browse listings
Close browser
- HAR saved to: ~/.reverse-api/runs/har/{id}/network.har
+ HAR saved to: ~/.reverse-api/runs/har/{run_id}/recording.har
Note: you can monitor browser requests with the Playwright MCP