From 175979fb1542c1adf797cbd16b1af5bd7dc6826e Mon Sep 17 00:00:00 2001 From: Farouk Adeleke Date: Thu, 22 Jan 2026 17:37:10 -0800 Subject: [PATCH] change signature --- config.json | 2 +- nanocode.py | 272 +-------------------------------------------------- program.json | 10 +- 3 files changed, 6 insertions(+), 278 deletions(-) diff --git a/config.json b/config.json index c150e42..3a011d0 100644 --- a/config.json +++ b/config.json @@ -4,7 +4,7 @@ "lm": "openrouter/openai/gpt-5.2-codex", "sub_lm": "openrouter/openai/gpt-5-mini", "api_base": "https://openrouter.ai/api/v1", - "max_tokens": 32000, + "max_tokens": 50000, "max_output_chars": 100000, "verbose": false, "track_usage": true diff --git a/nanocode.py b/nanocode.py index edacd59..0df9594 100644 --- a/nanocode.py +++ b/nanocode.py @@ -1,12 +1,8 @@ import os -import re import glob as globlib -import subprocess -import shlex -import json -import tempfile from modaic import PrecompiledProgram, PrecompiledConfig import dspy +import re # --- Modaic --- @@ -24,29 +20,6 @@ YELLOW = "\033[33m" RED = "\033[31m" MAGENTA = "\033[35m" -# --- Display utilities --- - -LONG_PASTE_THRESHOLD = int(os.environ.get("NANOCODE_LONG_PASTE_THRESHOLD", "4000")) - - -def save_long_paste(text: str) -> str: - fd, path = tempfile.mkstemp(prefix="nanocode_paste_", suffix=".txt") - with os.fdopen(fd, "w") as handle: - handle.write(text) - return path - - - -def separator(): - """Return a horizontal separator line that fits the terminal width.""" - return f"{DIM}{'─' * min(os.get_terminal_size().columns, 80)}{RESET}" - - -def render_markdown(text): - """Convert basic markdown bold syntax to ANSI bold.""" - return re.sub(r"\*\*(.+?)\*\*", f"{BOLD}\\1{RESET}", text) - - # --- File operations --- @@ -200,7 +173,6 @@ AVAILABLE_MODELS = { "8": ("Minimax M2.1", "minimax/minimax-m2.1"), } - def select_model(): """Interactive model selection or use environment variable.""" print(f"\n{BOLD}Select a model:{RESET}") @@ -243,16 +215,13 @@ class CodingAssistant(dspy.Signature): answer: str = dspy.OutputField( desc="Your response to the user after completing the task" ) - affected_files: list[str] = dspy.OutputField( - desc="List of files that were written or modified during the task" - ) class RLMCodingConfig(PrecompiledConfig): max_iters: int = 50 lm: str = "openrouter/openai/gpt-5.2-codex" sub_lm: str = "openrouter/openai/gpt-5-mini" api_base: str = "https://openrouter.ai/api/v1" - max_tokens: int = 32000 + max_tokens: int = 50000 max_output_chars: int = 100000 verbose: bool = False track_usage: bool = True @@ -328,243 +297,6 @@ class RLMCodingProgram(PrecompiledProgram): new_instance.set_lm(self.lm) self.agent = new_instance - -def main(): - model = select_model() - - # Add openrouter/ prefix if not already present - if not model.startswith("openrouter/"): - model = f"openrouter/{model}" - - config = RLMCodingConfig() - config.lm = model - - agent = RLMCodingProgram(config) - print( - f"{BOLD}NANOCODE DSPY{RESET} | {DIM}{agent.config.lm} | {os.getcwd()}{RESET}\n" - ) - - # Conversation history for context - history = [] - - # MCP servers registry - mcp_servers = {} - - def register_mcp_server(name, server): - tool_names = [] - for tool in server.tools: - tool_name = f"{name}_{tool.__name__}" - agent.set_tool(tool_name, tool) - tool_names.append(tool_name) - return tool_names - - while True: - try: - print(separator()) - user_input = input(f"{BOLD}{BLUE}❯{RESET} ").strip() - print(separator()) - - tmp_paste_path = None - if len(user_input) > LONG_PASTE_THRESHOLD: - tmp_paste_path = save_long_paste(user_input) - print( - f"{YELLOW}⏺ Long paste detected ({len(user_input)} chars). Saved to {tmp_paste_path}{RESET}" - ) - user_input = ( - f"The user pasted a long input ({len(user_input)} chars). " - f"It has been saved to {tmp_paste_path}. " - "Use read_file to view it. The file will be deleted after this response." - ) - - if not user_input: - continue - if user_input in ("/q", "exit"): - break - if user_input == "/c": - history = [] - print(f"{GREEN}⏺ Cleared conversation{RESET}") - continue - if user_input == "/model": - print(f"\n{BOLD}Current model: {agent.config.lm}{RESET}") - print(f"\n{BOLD}Select a new model:{RESET}") - for key, (name, model_id) in AVAILABLE_MODELS.items(): - print(f" {BLUE}{key}{RESET}. {name} ({DIM}{model_id}{RESET})") - print(f" {BLUE}c{RESET}. Custom model (enter manually)") - print(f" {BLUE}k{RESET}. Keep current model") - - choice = input(f"\n{BOLD}{BLUE}❯{RESET} Enter choice: ").strip().lower() - - if choice == "k": - print(f"{GREEN}⏺ Keeping current model: {agent.config.lm}{RESET}") - continue - elif choice in AVAILABLE_MODELS: - name, model_id = AVAILABLE_MODELS[choice] - new_model = ( - model_id - if model_id.startswith("openrouter/") - else f"openrouter/{model_id}" - ) - config.lm = new_model - agent = RLMCodingProgram(config) - for server_name, info in mcp_servers.items(): - info["tools"] = register_mcp_server(server_name, info["server"]) - print(f"{GREEN}⏺ Switched to: {name} ({new_model}){RESET}") - elif choice == "c": - custom_model = input( - f"{BOLD}{BLUE}❯{RESET} Enter model ID: " - ).strip() - if custom_model: - new_model = ( - custom_model - if custom_model.startswith("openrouter/") - else f"openrouter/{custom_model}" - ) - config.lm = new_model - agent = RLMCodingProgram(config) - for server_name, info in mcp_servers.items(): - info["tools"] = register_mcp_server( - server_name, info["server"] - ) - print(f"{GREEN}⏺ Switched to custom model: {new_model}{RESET}") - else: - print(f"{RED}⏺ Invalid model ID, keeping current model{RESET}") - else: - print(f"{RED}⏺ Invalid choice, keeping current model{RESET}") - continue - - if user_input.startswith("/add-mcp"): - parts = shlex.split(user_input) - args = parts[1:] - if not args: - print( - f"{YELLOW}⏺ Usage: /add-mcp [--auth |--oauth] [--headers ''] [--auto-auth|--no-auto-auth]{RESET}" - ) - continue - - name = None - auth = None - headers = None - auto_auth = None - positional = [] - i = 0 - while i < len(args): - if args[i] in ("--name", "-n") and i + 1 < len(args): - name = args[i + 1] - i += 2 - elif args[i].startswith("--auth="): - auth = args[i].split("=", 1)[1] - i += 1 - elif args[i] == "--auth" and i + 1 < len(args): - auth = args[i + 1] - i += 2 - elif args[i] == "--oauth": - auth = "oauth" - i += 1 - elif args[i] == "--auto-auth": - auto_auth = True - i += 1 - elif args[i] == "--no-auto-auth": - auto_auth = False - i += 1 - elif args[i].startswith("--headers="): - headers = json.loads(args[i].split("=", 1)[1]) - i += 1 - elif args[i] == "--headers" and i + 1 < len(args): - headers = json.loads(args[i + 1]) - i += 2 - else: - positional.append(args[i]) - i += 1 - - server_cmd = None - if positional: - if name is None and len(positional) >= 2: - name = positional[0] - server_cmd = " ".join(positional[1:]) - else: - server_cmd = " ".join(positional) - - if not server_cmd: - print( - f"{YELLOW}⏺ Usage: /add-mcp [--auth |--oauth] [--headers ''] [--auto-auth|--no-auto-auth]{RESET}" - ) - continue - - if not name: - name = re.sub(r"[^a-zA-Z0-9_]+", "_", server_cmd).strip("_") - if not name: - name = f"mcp_{len(mcp_servers) + 1}" - - if name in mcp_servers: - for tool_name in mcp_servers[name]["tools"]: - agent.remove_tool(tool_name) - - try: - from mcp2py import load - - kwargs = {} - if auth is not None: - kwargs["auth"] = auth - if headers: - kwargs["headers"] = headers - if auto_auth is not None: - kwargs["auto_auth"] = auto_auth - - server = load(server_cmd, **kwargs) - tool_names = register_mcp_server(name, server) - mcp_servers[name] = {"server": server, "tools": tool_names} - - print( - f"{GREEN}⏺ Added MCP server '{name}' with {len(tool_names)} tools{RESET}" - ) - print(f"{GREEN}⏺ Tools: {list(agent.tools.keys())}{RESET}") - except Exception as err: - print(f"{RED}⏺ Failed to add MCP server: {err}{RESET}") - - continue - - # Build context from history - context = f"Working directory: {os.getcwd()}\n" - if history: - context += "\nPrevious conversation:\n" - for h in history[-5:]: # Keep last 5 exchanges - context += f"User: {h['user']}\nAssistant: {h['assistant']}\n\n" - - task = f"{context}\nCurrent task: {user_input}" - - print(f"\n{CYAN}⏺{RESET} Thinking...", flush=True) - - # Run the RLM agent - try: - result = agent(task=task) - finally: - if tmp_paste_path: - try: - os.remove(tmp_paste_path) - except OSError: - pass - - # Display the answer - print(f"\n{CYAN}⏺{RESET} {render_markdown(result.answer)}") - - # Display usage - print(f"\n{MAGENTA}⏺ Debug Prediction: {result}{RESET}") - - # Save to history - history.append({"user": user_input, "assistant": result.answer}) - - print() - - except (KeyboardInterrupt, EOFError): - break - except Exception as err: - import traceback - - traceback.print_exc() - print(f"{RED}⏺ Error: {err}{RESET}") - - if __name__ == "__main__": agent = RLMCodingProgram(RLMCodingConfig()) agent.push_to_hub(MODAIC_REPO_PATH, commit_message="change signature", branch="dev") - #main() diff --git a/program.json b/program.json index bdd17b1..4ad808b 100644 --- a/program.json +++ b/program.json @@ -4,7 +4,7 @@ "train": [], "demos": [], "signature": { - "instructions": "You are a concise coding assistant with access to sub agents.\n\nYou are tasked with producing the following outputs given the inputs `task`:\n- {answer}\n- {affected_files} # note: the value you produce must adhere to the JSON schema: {\"type\": \"array\", \"items\": {\"type\": \"string\"}}\n\nYou have access to a Python REPL environment. Write Python code and it will be executed. You will see the output, then write more code based on what you learned. This is an iterative process.\n\nAvailable:\n- Variables: `task` (your input data)\n- `llm_query(prompt)` - query a sub-LLM (~500K char capacity) for semantic analysis\n- `llm_query_batched(prompts)` - query multiple prompts concurrently (much faster for multiple queries)\n- `print()` - ALWAYS print to see results\n- `SUBMIT(answer, affected_files)` - submit final output when done\n- Standard libraries: re, json, collections, math, etc.\n\nIMPORTANT: This is ITERATIVE. Each code block you write will execute, you'll see the output, then you decide what to do next. Do NOT try to solve everything in one step.\n\n1. EXPLORE FIRST - Look at your data before processing it. Print samples, check types/lengths, understand the structure.\n2. ITERATE - Write small code snippets, observe outputs, then decide next steps. State persists between iterations.\n3. VERIFY BEFORE SUBMITTING - If results seem wrong (zeros, empty, unexpected), reconsider your approach.\n4. USE llm_query FOR SEMANTICS - String matching finds WHERE things are; llm_query understands WHAT things mean.\n5. MINIMIZE RETYPING (INPUTS & OUTPUTS) - When values are long, precise, or error-prone (IDs, numbers, code, quotes), re-access them via variables and parse/compute in code instead of retyping. Use small, targeted prints to sanity-check, but avoid manual copying when variables can carry the exact value.\n6. SUBMIT ONLY AFTER SEEING OUTPUTS - SUBMIT ends the current run immediately. If you need to inspect printed output, run it in one step, review the result, then call SUBMIT in a later step.\n\nYou have max 50 sub-LLM calls. When done, call SUBMIT() with your output.\nAdditional tools available (use these instead of standard library equivalents):\n- `read_file(path: str, offset: int, limit: int) -> str` - Read file contents with line numbers.\n- `write_file(path: str, content: str) -> str` - Write content to a file.\n- `edit_file(path: str, old: str, new: str, replace_all: bool) -> str` - Replace text in a file.\n- `glob_files(pattern: str, path: str) -> str` - Find files matching a glob pattern, sorted by modification time.\n- `grep_files(pattern: str, path: str) -> str` - Search files for a regex pattern.\n- `run_bash(cmd: str) -> str` - Run a shell command and return output.", + "instructions": "You are a concise coding assistant with access to sub agents.\n\nYou are tasked with producing the following outputs given the inputs `task`:\n- {answer}\n\nYou have access to a Python REPL environment. Write Python code and it will be executed. You will see the output, then write more code based on what you learned. This is an iterative process.\n\nAvailable:\n- Variables: `task` (your input data)\n- `llm_query(prompt)` - query a sub-LLM (~500K char capacity) for semantic analysis\n- `llm_query_batched(prompts)` - query multiple prompts concurrently (much faster for multiple queries)\n- `print()` - ALWAYS print to see results\n- `SUBMIT(answer)` - submit final output when done\n- Standard libraries: re, json, collections, math, etc.\n\nIMPORTANT: This is ITERATIVE. Each code block you write will execute, you'll see the output, then you decide what to do next. Do NOT try to solve everything in one step.\n\n1. EXPLORE FIRST - Look at your data before processing it. Print samples, check types/lengths, understand the structure.\n2. ITERATE - Write small code snippets, observe outputs, then decide next steps. State persists between iterations.\n3. VERIFY BEFORE SUBMITTING - If results seem wrong (zeros, empty, unexpected), reconsider your approach.\n4. USE llm_query FOR SEMANTICS - String matching finds WHERE things are; llm_query understands WHAT things mean.\n5. MINIMIZE RETYPING (INPUTS & OUTPUTS) - When values are long, precise, or error-prone (IDs, numbers, code, quotes), re-access them via variables and parse/compute in code instead of retyping. Use small, targeted prints to sanity-check, but avoid manual copying when variables can carry the exact value.\n6. SUBMIT ONLY AFTER SEEING OUTPUTS - SUBMIT ends the current run immediately. If you need to inspect printed output, run it in one step, review the result, then call SUBMIT in a later step.\n\nYou have max 50 sub-LLM calls. When done, call SUBMIT() with your output.\nAdditional tools available (use these instead of standard library equivalents):\n- `read_file(path: str, offset: int, limit: int) -> str` - Read file contents with line numbers.\n- `write_file(path: str, content: str) -> str` - Write content to a file.\n- `edit_file(path: str, old: str, new: str, replace_all: bool) -> str` - Replace text in a file.\n- `glob_files(pattern: str, path: str) -> str` - Find files matching a glob pattern, sorted by modification time.\n- `grep_files(pattern: str, path: str) -> str` - Search files for a regex pattern.\n- `run_bash(cmd: str) -> str` - Run a shell command and return output.", "fields": [ { "prefix": "Variables Info:", @@ -37,7 +37,7 @@ "launch_kwargs": {}, "train_kwargs": {}, "temperature": null, - "max_tokens": 32000, + "max_tokens": 50000, "api_base": "https://openrouter.ai/api/v1", "track_usage": true } @@ -60,10 +60,6 @@ { "prefix": "Answer:", "description": "Your response to the user after completing the task" - }, - { - "prefix": "Affected Files:", - "description": "List of files that were written or modified during the task" } ] }, @@ -76,7 +72,7 @@ "launch_kwargs": {}, "train_kwargs": {}, "temperature": null, - "max_tokens": 32000, + "max_tokens": 50000, "api_base": "https://openrouter.ai/api/v1", "track_usage": true }