chore: release api v1.0.0

- Event-based streaming for task agent
- Retry logic when LLM responds without calling tools
- Hardened prompts to enforce tool use
- Working directory context in all agent prompts
- Project paused: local LLMs not capable enough for agentic use

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

webber-api/src/domains/agents/explore/agent.py

@@ -79,6 +79,14 @@ class ExploreAgentImpl(BaseAgent):
         from src.domains.agents.explore.tools import register_explore_tools
         register_explore_tools(agent)
 
+    def _build_prompt_with_context(self, prompt: str, working_dir: str) -> str:
+        """Build the prompt with working directory context."""
+        return f"""Working directory: {working_dir}
+
+Use paths within this working directory for file operations.
+
+User request: {prompt}"""
+
     @logged()
     async def run(
         self,
@@ -98,16 +106,20 @@ class ExploreAgentImpl(BaseAgent):
         Returns:
             Agent response with findings
         """
+        effective_working_dir = working_dir or os.getcwd()
+
         ctx = ExploreContext(
-            working_dir=working_dir or os.getcwd(),
+            working_dir=effective_working_dir,
             allowed_paths=allowed_paths or self._settings.effective_allowed_paths,
             timeout_seconds=self._settings.tool_timeout_seconds,
         )
 
+        full_prompt = self._build_prompt_with_context(prompt, effective_working_dir)
+
         async with trace_span("explore_agent_run"):
             try:
                 # Use run() not run_stream() - Ollama has bugs with streaming + tools
-                result = await self.agent.run(prompt, deps=ctx)
+                result = await self.agent.run(full_prompt, deps=ctx)
                 return result.output
             except Exception as e:
                 logger.exception(f"Explore agent error: {e}")
@@ -126,15 +138,19 @@ class ExploreAgentImpl(BaseAgent):
 
         Yields text chunks as they become available.
         """
+        effective_working_dir = working_dir or os.getcwd()
+
         ctx = ExploreContext(
-            working_dir=working_dir or os.getcwd(),
+            working_dir=effective_working_dir,
             allowed_paths=allowed_paths or self._settings.effective_allowed_paths,
             timeout_seconds=self._settings.tool_timeout_seconds,
         )
 
+        full_prompt = self._build_prompt_with_context(prompt, effective_working_dir)
+
         async with trace_span("explore_agent_stream"):
             try:
-                async with self.agent.run_stream(prompt, deps=ctx) as result:
+                async with self.agent.run_stream(full_prompt, deps=ctx) as result:
                     async for chunk in result.stream_text():
                         yield chunk
             except Exception as e: