claude-fsd 1.3.4 โ†’ 1.3.5

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -1,19 +1,28 @@
1
1
  # claude-fsd
2
2
 
3
- Claude Full Self Drive (FSD) - Your AI-powered development team on autopilot.
3
+ Automated project development - let AI build your project while you sleep.
4
4
 
5
5
  ## What is this?
6
6
 
7
- Remember when junior developers were going to be replaced by AI? Well, the tables have turned. As explained in the excellent article ["Revenge of the Junior Developer"](https://sourcegraph.com/blog/revenge-of-the-junior-developer), AI has actually made junior developers more powerful than ever.
7
+ claude-fsd is an automated development system that runs continuous development cycles without constant supervision. Write a brief description of what you want, answer some clarifying questions, then let it build your project autonomously.
8
8
 
9
- This tool takes that concept to the next level by creating an entire **agent fleet** - multiple AI agents working together like a development team:
9
+ Think of it like **Tesla FSD for code** - it handles the driving, but you should keep an eye on it and occasionally take the wheel when needed.
10
+
11
+ ### How the Agent System Works
12
+
13
+ The system operates with multiple specialized AI agents working together:
10
14
 
11
15
  - ๐Ÿง‘โ€๐Ÿ’ป **Developer Agent**: Writes code, implements features, fixes bugs
12
- - ๐Ÿ“‹ **Planner Agent**: Breaks down tasks, manages the development roadmap
16
+ - ๐Ÿ“‹ **Planner Agent**: Breaks down tasks, manages the development roadmap
13
17
  - ๐Ÿ‘€ **Reviewer Agent**: Reviews code quality, catches issues, ensures best practices
14
18
  - ๐Ÿงช **Tester Agent**: Runs tests, validates changes, commits clean code
15
19
 
16
- Think of it as having a full development team on autopilot - like Tesla's Full Self-Driving, but for code! The AI agents work in cycles, planning tasks, implementing them, reviewing the work, and then moving on to the next task, with minimal human intervention.
20
+ The agents work in continuous cycles:
21
+ ```
22
+ Plan โ†’ Develop โ†’ Review โ†’ Test โ†’ Commit โ†’ Repeat
23
+ ```
24
+
25
+ You can leave it running while you grab lunch or sleep - it just keeps going until it thinks the project is complete.
17
26
 
18
27
  ## Installation
19
28
 
@@ -23,12 +32,15 @@ npm install -g claude-fsd
23
32
 
24
33
  ## Quick Start
25
34
 
26
- Just run:
27
35
  ```bash
28
- claude-fsd # or claudefsd - both work the same
36
+ claude-fsd
29
37
  ```
30
38
 
31
- You'll get an interactive menu to choose what you want to do. It's that simple!
39
+ The system will walk you through the setup process:
40
+
41
+ 1. **Create a BRIEF.md** - Write a short description of what you want to build
42
+ 2. **Answer questions in QUESTIONS.md** - The AI will generate clarifying questions for you to answer
43
+ 3. **Let it rip** - Start the automated development process
32
44
 
33
45
  ## Commands
34
46
 
@@ -71,17 +83,34 @@ Generates an initial project plan from scratch based on:
71
83
 
72
84
  ## How it Works
73
85
 
74
- 1. **You define what you want** in `BRIEF.md`
75
- 2. **AI analyst creates a series of questions** in `docs/QUESTIONS.md` for you to answer
76
- 3. **You answer the questions in the file**
77
- 4. **AI architect creates detailed notes and plan** in `docs/CLAUDE-NOTES.md` and `docs/PLAN.md`
78
- 5. **The agent fleet executes** the plan task by task
79
- 6. **You review and guide** the process when needed, in a separate `claude` instance, or just by watching the output and reading the files
86
+ 1. **Write a BRIEF.md** - Describe what you want to build
87
+ 2. **Answer AI-generated questions** in `QUESTIONS.md`
88
+ 3. **Start the development loop** - The system automatically:
89
+ - Picks the next task from your plan
90
+ - Implements the feature
91
+ - Reviews the code
92
+ - Runs tests and commits changes
93
+ - Repeats until complete
80
94
 
81
- The agents work in a continuous loop:
82
- ```
83
- Plan โ†’ Develop โ†’ Review โ†’ Test โ†’ Commit โ†’ Repeat
84
- ```
95
+ ## Monitoring Progress (Like Tesla FSD)
96
+
97
+ This isn't sci-fi level "sleep through the entire project" automation - it's more like Tesla FSD. The system handles most of the work, but you should monitor it and be ready to intervene when needed.
98
+
99
+ **Recommended monitoring approach:**
100
+ - **Run a parallel Claude session** - Open another Claude window/tab to chat about the project
101
+ - **Check status periodically** - Ask the parallel Claude: "What's the current status of my project?"
102
+ - **Review the plan** - Look at `docs/PLAN.md` to see what's been completed and what's next
103
+ - **Watch for drift** - If the system goes off track, intervene before it gets too far
104
+
105
+ **When you need to course-correct:**
106
+ - **Update the plan** - Add urgent fixes or redirections to the top of `docs/PLAN.md`
107
+ - **Direct intervention** - Use your parallel Claude session to directly fix issues
108
+ - **The system adapts** - claude-fsd will pick up plan changes on the next loop iteration
109
+
110
+ **Interruptible design:**
111
+ - Hit Ctrl+C anytime to pause
112
+ - Restart later with `claude-fsd dev`
113
+ - Perfect for running overnight, during meetings, or while getting lunch
85
114
 
86
115
  ## Requirements
87
116
 
@@ -110,11 +139,12 @@ your-project/
110
139
 
111
140
  ## Tips for Success
112
141
 
113
- 1. **Start small** - Break down your project into small, clear tasks
114
- 2. **Be specific** - The clearer your requirements, the better the results
115
- 3. **Review regularly** - Check in on what the agents are doing
116
- 4. **Use version control** - The agents will commit changes, but you should review them
117
- 5. **Trust but verify** - The agents are good but not perfect
142
+ 1. **Keep your BRIEF.md concise** - A few clear paragraphs work better than lengthy specifications
143
+ 2. **Answer questions thoroughly** - The AI's questions help it understand your exact needs
144
+ 3. **Monitor periodically** - Check progress while it runs, especially during initial cycles
145
+ 4. **Use the plan as your steering wheel** - Update `docs/PLAN.md` to guide development direction
146
+ 5. **Trust the process** - Let it run autonomously, but verify the results
147
+
118
148
 
119
149
  ## License
120
150
 
package/bin/claudefsd-dev CHANGED
@@ -8,7 +8,14 @@ $(dirname "$0")/claudefsd-check-dependencies
8
8
  # Add counter for loop iterations
9
9
  LOOP_COUNTER=0
10
10
 
11
+ # Failure detection variables
12
+ CONSECUTIVE_FAST_ITERATIONS=0
13
+ MIN_ITERATION_TIME=300 # 5 minutes in seconds
14
+
11
15
  while true; do
16
+ # Record iteration start time
17
+ ITERATION_START_TIME=$(date +%s)
18
+
12
19
  # Increment loop counter
13
20
  LOOP_COUNTER=$((LOOP_COUNTER + 1))
14
21
 
@@ -233,6 +240,44 @@ If you have any ideas for the future, you can add them to IDEAS.md.
233
240
  fi
234
241
  set -e
235
242
 
243
+ # Calculate iteration duration and check for failure patterns
244
+ ITERATION_END_TIME=$(date +%s)
245
+ ITERATION_DURATION=$((ITERATION_END_TIME - ITERATION_START_TIME))
246
+
247
+ echo -e "\033[36mIteration $LOOP_COUNTER completed in ${ITERATION_DURATION}s\033[0m"
248
+
249
+ # Check if iteration was suspiciously fast (likely failure mode)
250
+ if [ $ITERATION_DURATION -lt $MIN_ITERATION_TIME ]; then
251
+ CONSECUTIVE_FAST_ITERATIONS=$((CONSECUTIVE_FAST_ITERATIONS + 1))
252
+ echo -e "\033[33mWarning: Fast iteration detected (${ITERATION_DURATION}s < ${MIN_ITERATION_TIME}s threshold)\033[0m"
253
+ echo -e "\033[33mConsecutive fast iterations: $CONSECUTIVE_FAST_ITERATIONS/3\033[0m"
254
+
255
+ # Exit if too many consecutive fast iterations (likely Claude API failure)
256
+ if [ $CONSECUTIVE_FAST_ITERATIONS -ge 3 ]; then
257
+ echo -e "\033[31m==================================================================\033[0m"
258
+ echo -e "\033[31m== FAILURE MODE DETECTED - THROTTLING ACTIVATED\033[0m"
259
+ echo -e "\033[31m==================================================================\033[0m"
260
+ echo -e "\033[31mDetected 3 consecutive iterations under ${MIN_ITERATION_TIME}s each.\033[0m"
261
+ echo -e "\033[31mThis usually indicates Claude API issues (token limits, etc).\033[0m"
262
+ echo -e "\033[31m\033[0m"
263
+ echo -e "\033[31mSuggested actions:\033[0m"
264
+ echo -e "\033[31m- Check your Claude API token limits\033[0m"
265
+ echo -e "\033[31m- Wait a few minutes and restart with: claude-fsd dev\033[0m"
266
+ echo -e "\033[31m- Review logs in: logs/\033[0m"
267
+ echo -e "\033[31m==================================================================\033[0m"
268
+ exit 1
269
+ fi
270
+
271
+ # Add exponential backoff delay for fast iterations
272
+ BACKOFF_DELAY=$((CONSECUTIVE_FAST_ITERATIONS * 60)) # 1min, 2min, 3min
273
+ echo -e "\033[33mApplying backoff delay: ${BACKOFF_DELAY}s\033[0m"
274
+ sleep $BACKOFF_DELAY
275
+ else
276
+ # Reset counter on successful iteration
277
+ CONSECUTIVE_FAST_ITERATIONS=0
278
+ echo -e "\033[32mNormal iteration timing - continuing...\033[0m"
279
+ fi
280
+
236
281
  sleep 1
237
282
  done
238
283
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "claude-fsd",
3
- "version": "1.3.4",
3
+ "version": "1.3.5",
4
4
  "description": "Claude Full Self Drive tools for autonomous AI-powered development",
5
5
  "bin": {
6
6
  "claude-fsd": "bin/claude-fsd",