rlm-code 0.1.6__tar.gz → 0.1.7__tar.gz

{rlm_code-0.1.6 → rlm_code-0.1.7}/CHANGELOG.md

@@ -5,6 +5,19 @@ All notable changes to this project are documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.1.7] - 2026-04-30
+
+### Added
+- HALO-style `trace_analysis` RLM environment for diagnosing agent harness failures from one-span-per-line JSONL traces.
+- Trace sidecar indexing with dataset rollups for trace counts, span counts, error traces, services, models, agents, token totals, and sample trace ids.
+- Bounded trace inspection actions: `get_dataset_overview`, `query_traces`, `count_traces`, `view_trace`, `search_trace`, and `view_spans`.
+- Large-trace safeguards: per-attribute truncation, oversized trace summaries, and higher-cap selected-span reads.
+- Tests for trace indexing, querying, searching, selected-span viewing, and trace environment actions.
+- Trace analysis documentation under the Core Engine docs.
+
+### Changed
+- `/rlm` command help now advertises `env=trace_analysis` for run, chat, and doctor workflows.
+
 ## [0.1.6] - 2026-02-20
 
 ### Added
@@ -56,3 +69,4 @@ Initial public release of **RLM Code**.
 
 [0.1.5]: https://github.com/SuperagenticAI/rlm-code/releases/tag/v0.1.5
 [0.1.6]: https://github.com/SuperagenticAI/rlm-code/releases/tag/v0.1.6
+[0.1.7]: https://github.com/SuperagenticAI/rlm-code/releases/tag/v0.1.7

{rlm_code-0.1.6 → rlm_code-0.1.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rlm-code
-Version: 0.1.6
+Version: 0.1.7
 Summary: RLM Code: Research Playground & Evaluation OS for Recursive Language Model Agentic Systems
 Project-URL: Homepage, https://github.com/SuperagenticAI/rlm-code
 Project-URL: Documentation, https://superagenticai.github.io/rlm-code/
@@ -118,20 +118,20 @@ RLM Code implements the [Recursive Language Models](https://arxiv.org/abs/2502.0
 
 RLM Code wraps this algorithm in an interactive terminal UI with built-in benchmarks, trajectory replay, and observability.
 
-## Release v0.1.6
+## Release v0.1.7
 
-This release adds the new CodeMode path as an opt-in harness strategy.
+This release adds HALO-style trace analysis as a new RLM environment.
 
-- New harness strategy: `strategy=codemode` (default remains `strategy=tool_call`)
-- MCP bridge flow for CodeMode: `search_tools` -> typed tool surface -> `call_tool_chain`
-- Guardrails before execution: blocked API classes plus timeout/size/tool-call caps
-- Benchmark telemetry for side-by-side comparison: `tool_call` vs `codemode`
-- Dedicated docs section for CodeMode: quickstart, architecture, guardrails, evaluation
+- New `trace_analysis` environment for diagnosing agent harness failures from OTel-shaped JSONL traces
+- Sidecar trace indexing with dataset overview, query, count, search, full-trace view, and selected-span view actions
+- Bounded payload handling for large traces, including oversized summaries and higher-cap surgical span reads
+- `/rlm` help/docs updated for `env=trace_analysis`
+- Dedicated trace analysis docs under the Core Engine section
 
 Example:
 
 ```text
-/harness run "implement feature and add tests" steps=8 mcp=on strategy=codemode mcp_server=codemode
+/rlm run "Find systemic harness failures trace=./traces.jsonl" env=trace_analysis steps=6
 ```
 
 ## Documentation
@@ -287,6 +287,62 @@ Notes:
 - In Local/BYOK connection modes, likely coding prompts in chat can auto-route to harness.
 - In ACP mode, auto-routing is intentionally off; use `/harness run ...` explicitly.
 
+### 8. CodeMode with UTCP and Cloudflare MCP
+
+Use these server entries in your project `rlm_config.yaml`:
+
+```yaml
+mcp_servers:
+  utcp-codemode:
+    name: utcp-codemode
+    description: "Local CodeMode MCP bridge"
+    enabled: true
+    auto_connect: false
+    timeout_seconds: 30
+    retry_attempts: 3
+    transport:
+      type: stdio
+      command: npx
+      args:
+        - "@utcp/code-mode-mcp"
+
+  cloudflare-codemode:
+    name: cloudflare-codemode
+    description: "Cloudflare MCP via remote bridge"
+    enabled: true
+    auto_connect: false
+    timeout_seconds: 30
+    retry_attempts: 3
+    transport:
+      type: stdio
+      command: npx
+      args:
+        - "mcp-remote"
+        - "https://mcp.cloudflare.com/mcp"
+```
+
+UTCP path (native CodeMode in current release):
+
+```text
+/mcp-connect utcp-codemode
+/mcp-tools utcp-codemode
+/harness run "analyze this repo, find TODO/FIXME, and create report.json" steps=3 mcp=on strategy=codemode mcp_server=utcp-codemode
+```
+
+Cloudflare path (recommended strategy today):
+
+```text
+/mcp-connect cloudflare-codemode
+/mcp-tools cloudflare-codemode
+/harness run "list available tools and run one safe read-only action, then summarize in 3 bullets" steps=3 mcp=on strategy=tool_call mcp_server=cloudflare-codemode
+```
+
+Notes:
+
+- On first Cloudflare connect, `mcp-remote` may ask for interactive authentication.
+- In this release, `strategy=codemode` expects the `search_tools` + `call_tool_chain` bridge contract.
+- If a remote MCP server exposes a different tool contract, use `strategy=tool_call`.
+
 ## How the RLM Loop Works
 
 Traditional LLM usage: paste your document into the prompt, ask a question, hope the model doesn't lose details in the middle.

{rlm_code-0.1.6 → rlm_code-0.1.7}/README.md

@@ -25,20 +25,20 @@ RLM Code implements the [Recursive Language Models](https://arxiv.org/abs/2502.0
 
 RLM Code wraps this algorithm in an interactive terminal UI with built-in benchmarks, trajectory replay, and observability.
 
-## Release v0.1.6
+## Release v0.1.7
 
-This release adds the new CodeMode path as an opt-in harness strategy.
+This release adds HALO-style trace analysis as a new RLM environment.
 
-- New harness strategy: `strategy=codemode` (default remains `strategy=tool_call`)
-- MCP bridge flow for CodeMode: `search_tools` -> typed tool surface -> `call_tool_chain`
-- Guardrails before execution: blocked API classes plus timeout/size/tool-call caps
-- Benchmark telemetry for side-by-side comparison: `tool_call` vs `codemode`
-- Dedicated docs section for CodeMode: quickstart, architecture, guardrails, evaluation
+- New `trace_analysis` environment for diagnosing agent harness failures from OTel-shaped JSONL traces
+- Sidecar trace indexing with dataset overview, query, count, search, full-trace view, and selected-span view actions
+- Bounded payload handling for large traces, including oversized summaries and higher-cap surgical span reads
+- `/rlm` help/docs updated for `env=trace_analysis`
+- Dedicated trace analysis docs under the Core Engine section
 
 Example:
 
 ```text
-/harness run "implement feature and add tests" steps=8 mcp=on strategy=codemode mcp_server=codemode
+/rlm run "Find systemic harness failures trace=./traces.jsonl" env=trace_analysis steps=6
 ```
 
 ## Documentation
@@ -194,6 +194,62 @@ Notes:
 - In Local/BYOK connection modes, likely coding prompts in chat can auto-route to harness.
 - In ACP mode, auto-routing is intentionally off; use `/harness run ...` explicitly.
 
+### 8. CodeMode with UTCP and Cloudflare MCP
+
+Use these server entries in your project `rlm_config.yaml`:
+
+```yaml
+mcp_servers:
+  utcp-codemode:
+    name: utcp-codemode
+    description: "Local CodeMode MCP bridge"
+    enabled: true
+    auto_connect: false
+    timeout_seconds: 30
+    retry_attempts: 3
+    transport:
+      type: stdio
+      command: npx
+      args:
+        - "@utcp/code-mode-mcp"
+
+  cloudflare-codemode:
+    name: cloudflare-codemode
+    description: "Cloudflare MCP via remote bridge"
+    enabled: true
+    auto_connect: false
+    timeout_seconds: 30
+    retry_attempts: 3
+    transport:
+      type: stdio
+      command: npx
+      args:
+        - "mcp-remote"
+        - "https://mcp.cloudflare.com/mcp"
+```
+
+UTCP path (native CodeMode in current release):
+
+```text
+/mcp-connect utcp-codemode
+/mcp-tools utcp-codemode
+/harness run "analyze this repo, find TODO/FIXME, and create report.json" steps=3 mcp=on strategy=codemode mcp_server=utcp-codemode
+```
+
+Cloudflare path (recommended strategy today):
+
+```text
+/mcp-connect cloudflare-codemode
+/mcp-tools cloudflare-codemode
+/harness run "list available tools and run one safe read-only action, then summarize in 3 bullets" steps=3 mcp=on strategy=tool_call mcp_server=cloudflare-codemode
+```
+
+Notes:
+
+- On first Cloudflare connect, `mcp-remote` may ask for interactive authentication.
+- In this release, `strategy=codemode` expects the `search_tools` + `call_tool_chain` bridge contract.
+- If a remote MCP server exposes a different tool contract, use `strategy=tool_call`.
+
 ## How the RLM Loop Works
 
 Traditional LLM usage: paste your document into the prompt, ask a question, hope the model doesn't lose details in the middle.

{rlm_code-0.1.6 → rlm_code-0.1.7}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "rlm-code"
-version = "0.1.6"
+version = "0.1.7"
 description = "RLM Code: Research Playground & Evaluation OS for Recursive Language Model Agentic Systems"
 readme = "README.md"
 license = "Apache-2.0"

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/__init__.py

@@ -5,5 +5,5 @@ This package provides tools for creating, managing, and optimizing DSPy componen
 through natural language interactions.
 """
 
-__version__ = "0.1.6"
+__version__ = "0.1.7"
 __author__ = "Super Agentic AI"

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/commands/slash_commands.py

@@ -1684,7 +1684,7 @@ class SlashCommandHandler:
         Manage RLM runs.
 
         Usage:
-            /rlm run <task> [steps=N] [timeout=N] [branch=N] [depth=N] [children=N] [parallel=N] [budget=N] [framework=<see /rlm frameworks>] [env=generic|dspy|pure_rlm] [sub=provider/model]
+            /rlm run <task> [steps=N] [timeout=N] [branch=N] [depth=N] [children=N] [parallel=N] [budget=N] [framework=<see /rlm frameworks>] [env=generic|dspy|pure_rlm|trace_analysis] [sub=provider/model]
             /rlm bench [list|preset=name] [mode=native|harness|direct-llm] [strategy=tool_call|codemode] [mcp=on|off] [mcp_server=name] [pack=path[,path2]] [limit=N] [steps=N] [timeout=N] [branch=N] [framework=<see /rlm frameworks>] [env=generic|dspy|pure_rlm] [sub=provider/model]
             /rlm bench compare [candidate=<id|path|latest>] [baseline=<id|path|previous>] [min_reward_delta=N] [min_completion_delta=N] [max_steps_increase=N]
             /rlm bench validate [candidate=<id|path|latest>] [baseline=<id|path|previous>] [min_reward_delta=N] [min_completion_delta=N] [max_steps_increase=N] [--json]
@@ -1696,8 +1696,8 @@ class SlashCommandHandler:
             /rlm status [run_id]
             /rlm abort [run_id|all]
             /rlm replay [run_id|latest]
-            /rlm doctor [env=generic|dspy|pure_rlm] [--json]
-            /rlm chat <message> [session=name] [env=generic|dspy|pure_rlm] [branch=N] [depth=N] [children=N] [parallel=N] [budget=N] [framework=<see /rlm frameworks>] [sub=provider/model]
+            /rlm doctor [env=generic|dspy|pure_rlm|trace_analysis] [--json]
+            /rlm chat <message> [session=name] [env=generic|dspy|pure_rlm|trace_analysis] [branch=N] [depth=N] [children=N] [parallel=N] [budget=N] [framework=<see /rlm frameworks>] [sub=provider/model]
             /rlm chat status [session=name]
             /rlm chat reset [session=name]
             /rlm observability
@@ -1708,14 +1708,14 @@ class SlashCommandHandler:
             console.print("[bold cyan]🧠 RLM Commands[/bold cyan]")
             console.print(
                 "  [yellow]/rlm run <task> [steps=N] [timeout=N] [branch=N] [depth=N] [children=N] "
-                f"[parallel=N] [budget=N] [framework={framework_opts}] [env=generic|dspy|pure_rlm] "
+                f"[parallel=N] [budget=N] [framework={framework_opts}] [env=generic|dspy|pure_rlm|trace_analysis] "
                 "[sub=provider/model][/yellow]"
             )
             console.print(
                 "  [yellow]/rlm bench [list|preset=name] [mode=native|harness|direct-llm] "
                 "[strategy=tool_call|codemode] [mcp=on|off] [mcp_server=name] "
                 "[pack=path[,path2]] [limit=N] [steps=N] "
-                f"[timeout=N] [branch=N] [framework={framework_opts}] [env=generic|dspy|pure_rlm] [sub=provider/model][/yellow]"
+                f"[timeout=N] [branch=N] [framework={framework_opts}] [env=generic|dspy|pure_rlm|trace_analysis] [sub=provider/model][/yellow]"
             )
             console.print(
                 "  [yellow]/rlm bench compare [candidate=<id|path|latest>] [baseline=<id|path|previous>] "
@@ -1741,9 +1741,9 @@ class SlashCommandHandler:
             console.print("  [yellow]/rlm status [run_id][/yellow]")
             console.print("  [yellow]/rlm abort [run_id|all][/yellow]")
             console.print("  [yellow]/rlm replay [run_id|latest][/yellow]")
-            console.print("  [yellow]/rlm doctor [env=generic|dspy|pure_rlm] [--json][/yellow]")
+            console.print("  [yellow]/rlm doctor [env=generic|dspy|pure_rlm|trace_analysis] [--json][/yellow]")
             console.print(
-                "  [yellow]/rlm chat <message> [session=name] [env=generic|dspy|pure_rlm] [branch=N] [depth=N] "
+                "  [yellow]/rlm chat <message> [session=name] [env=generic|dspy|pure_rlm|trace_analysis] [branch=N] [depth=N] "
                 f"[children=N] [parallel=N] [budget=N] [framework={framework_opts}] "
                 "[sub=provider/model][/yellow]"
             )
@@ -2135,7 +2135,7 @@ class SlashCommandHandler:
             task = " ".join(task_tokens).strip()
             if not task:
                 show_error_message(
-                    "Usage: /rlm run <task> [steps=N] [timeout=N] [env=generic|dspy|pure_rlm] "
+                    "Usage: /rlm run <task> [steps=N] [timeout=N] [env=generic|dspy|pure_rlm|trace_analysis] "
                     "[depth=N] [children=N] [parallel=N] [budget=N] "
                     f"[framework={framework_opts}] "
                     "[branch=N] [sub=provider/model]"

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/mcp/__init__.py

@@ -17,7 +17,7 @@ from .exceptions import (
 )
 from .session_wrapper import MCPSessionWrapper
 
-__version__ = "0.1.6"
+__version__ = "0.1.7"
 
 __all__ = [
     "MCPClientManager",

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/rlm/action_planner.py

@@ -15,6 +15,7 @@ from .environments import (
     DSPyCodingRLMEnvironment,
     GenericRLMEnvironment,
     RLMEnvironment,
+    TraceAnalysisEnvironment,
 )
 from .pure_rlm_environment import PureRLMConfig, PureRLMEnvironment
 
@@ -276,6 +277,8 @@ class ActionPlannerMixin:
             )
         if isinstance(env, DSPyCodingRLMEnvironment):
             return DSPyCodingRLMEnvironment(workdir=workdir, reward_profile=self.reward_profile)
+        if isinstance(env, TraceAnalysisEnvironment):
+            return TraceAnalysisEnvironment(workdir=workdir, reward_profile=self.reward_profile)
         if isinstance(env, GenericRLMEnvironment):
             return GenericRLMEnvironment(workdir=workdir, reward_profile=self.reward_profile)
         # Fallback to generic environment in preview if an unknown env type appears.

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/rlm/environments.py

@@ -286,6 +286,251 @@ class GenericRLMEnvironment:
         return "Execution failed without stderr."
 
 
+class TraceAnalysisEnvironment(GenericRLMEnvironment):
+    """HALO-style trace analysis environment over one-span-per-line JSONL traces."""
+
+    name = "trace_analysis"
+
+    def __init__(
+        self,
+        workdir: Path | None = None,
+        reward_profile: RLMRewardProfile | dict[str, Any] | None = None,
+    ):
+        super().__init__(workdir=workdir, reward_profile=reward_profile)
+        self._trace_path: Path | None = None
+        self._store: Any | None = None
+
+    def system_prompt(self) -> str:
+        return (
+            "You are an RLM planner specialized for analyzing agent execution traces.\n"
+            "Return ONLY valid JSON object with keys:\n"
+            "{"
+            '"action": "set_trace_path" | "get_dataset_overview" | "query_traces" | '
+            '"count_traces" | "view_trace" | "search_trace" | "view_spans" | "final", '
+            '"trace_path": "<path to JSONL traces>", '
+            '"filters": {"has_errors": true, "model_names": ["..."], "service_names": ["..."], '
+            '"agent_names": ["..."], "project_id": "..."}, '
+            '"trace_id": "<trace id>", '
+            '"span_ids": ["<span id>"], '
+            '"pattern": "<literal substring>", '
+            '"limit": <integer>, '
+            '"offset": <integer>, '
+            '"rationale": "<brief reason>", '
+            '"done": true|false, '
+            '"final_response": "<required when action=final>"'
+            "}\n"
+            "Rules:\n"
+            "- Load a trace file first if one is not already active.\n"
+            "- Always begin analysis with get_dataset_overview.\n"
+            "- Use query_traces to choose real trace ids; never invent trace ids.\n"
+            "- For large traces, prefer search_trace followed by view_spans.\n"
+            "- Identify systemic harness failures, not one-off anomalies.\n"
+            "- Output JSON only."
+        )
+
+    def planner_prompt(
+        self, task: str, memory: list[str], trajectory: list[dict[str, Any]], step_index: int
+    ) -> str:
+        inferred = self._extract_trace_path(task)
+        if inferred is not None and inferred != self._trace_path:
+            try:
+                self._load_store(inferred)
+            except Exception:
+                # Surface the failure through the prompt; execute_action will return
+                # the structured error if the planner attempts to use the path.
+                self._trace_path = inferred
+                self._store = None
+
+        base = super().planner_prompt(task, memory, trajectory, step_index)
+        active = str(self._trace_path) if self._trace_path is not None else "(none)"
+        overview = ""
+        if self._store is not None:
+            try:
+                data = self._store.get_overview({})
+                overview = (
+                    f"\nActive trace overview: traces={data['total_traces']} "
+                    f"spans={data['total_spans']} errors={data['error_trace_count']} "
+                    f"sample_trace_ids={data['sample_trace_ids'][:5]}"
+                )
+            except Exception:
+                overview = ""
+        return (
+            f"{base}\n\n"
+            f"Trace analysis environment.\n"
+            f"Active trace path: {active}\n"
+            "If the task includes trace=<path> or trace_path=<path>, use that file.\n"
+            "Goal: produce a concise evidence report of repeated harness failure modes "
+            "with concrete trace ids/spans and suggested harness changes."
+            f"{overview}"
+        )
+
+    def execute_action(
+        self,
+        action: dict[str, Any],
+        execution_engine: Any,
+        exec_timeout: int,
+        llm_connector: Any | None = None,
+    ) -> EnvironmentActionResult:
+        action_name = str(action.get("action", "")).strip().lower()
+        if action_name == "final":
+            return super().execute_action(
+                action,
+                execution_engine,
+                exec_timeout,
+                llm_connector=llm_connector,
+            )
+
+        try:
+            if action_name == "set_trace_path":
+                store = self._store_from_action(action, required_path=True)
+                return EnvironmentActionResult(
+                    observation={
+                        "success": True,
+                        "trace_path": str(store.trace_path),
+                        "index_path": str(store.index_path),
+                        "overview": store.get_overview({}),
+                    },
+                    reward=0.55,
+                    memory_note=f"Loaded trace dataset: {store.trace_path}",
+                )
+
+            store = self._store_from_action(action, required_path=False)
+            filters = action.get("filters") if isinstance(action.get("filters"), dict) else {}
+
+            if action_name == "get_dataset_overview":
+                return self._ok(
+                    observation=store.get_overview(filters),
+                    reward=0.45,
+                    memory_note="Loaded trace dataset overview.",
+                )
+            if action_name == "query_traces":
+                return self._ok(
+                    observation=store.query_traces(
+                        filters,
+                        limit=self._int_arg(action, "limit", 50, minimum=1, maximum=200),
+                        offset=self._int_arg(action, "offset", 0, minimum=0, maximum=1_000_000),
+                    ),
+                    reward=0.5,
+                    memory_note="Queried trace summaries.",
+                )
+            if action_name == "count_traces":
+                return self._ok(
+                    observation=store.count_traces(filters),
+                    reward=0.35,
+                    memory_note="Counted traces matching filters.",
+                )
+            if action_name == "view_trace":
+                trace_id = self._required_str(action, "trace_id")
+                return self._ok(
+                    observation=store.view_trace(trace_id),
+                    reward=0.65,
+                    memory_note=f"Viewed trace {trace_id}.",
+                )
+            if action_name == "search_trace":
+                trace_id = self._required_str(action, "trace_id")
+                pattern = self._required_str(action, "pattern")
+                return self._ok(
+                    observation=store.search_trace(
+                        trace_id,
+                        pattern,
+                        limit=self._int_arg(action, "limit", 100, minimum=1, maximum=500),
+                    ),
+                    reward=0.65,
+                    memory_note=f"Searched trace {trace_id} for {pattern!r}.",
+                )
+            if action_name == "view_spans":
+                trace_id = self._required_str(action, "trace_id")
+                span_ids = action.get("span_ids")
+                if not isinstance(span_ids, list) or not span_ids:
+                    raise ValueError("view_spans requires non-empty span_ids list")
+                return self._ok(
+                    observation=store.view_spans(trace_id, [str(item) for item in span_ids]),
+                    reward=0.7,
+                    memory_note=f"Viewed selected spans for trace {trace_id}.",
+                )
+        except Exception as exc:
+            return EnvironmentActionResult(
+                observation={"success": False, "error": f"{type(exc).__name__}: {exc}"},
+                reward=-0.25,
+                memory_note=f"Trace analysis action failed: {type(exc).__name__}.",
+            )
+
+        return EnvironmentActionResult(
+            observation={"success": False, "error": f"Unsupported action '{action_name}'."},
+            reward=-0.2,
+            memory_note="Planner produced unsupported trace action.",
+        )
+
+    def doctor_checks(self) -> list[EnvironmentDoctorCheck]:
+        checks = super().doctor_checks()
+        checks.append(
+            EnvironmentDoctorCheck(
+                name="trace_analysis",
+                status="pass",
+                detail="Trace analysis environment is available.",
+            )
+        )
+        return checks
+
+    def _ok(self, *, observation: dict[str, Any], reward: float, memory_note: str) -> EnvironmentActionResult:
+        payload = {"success": True, **observation}
+        return EnvironmentActionResult(
+            observation=payload,
+            reward=reward,
+            memory_note=memory_note,
+        )
+
+    def _store_from_action(self, action: dict[str, Any], *, required_path: bool):
+        raw = action.get("trace_path") or action.get("path")
+        if isinstance(raw, str) and raw.strip():
+            return self._load_store(Path(raw.strip()).expanduser())
+        if self._store is not None:
+            return self._store
+        if required_path:
+            raise ValueError("trace_path is required")
+        raise ValueError("no trace dataset loaded; pass trace_path or use set_trace_path first")
+
+    def _load_store(self, trace_path: Path):
+        from ..traces import TraceStore
+
+        resolved = trace_path if trace_path.is_absolute() else (self.workdir / trace_path)
+        store = TraceStore.load(resolved)
+        self._trace_path = resolved.resolve()
+        self._store = store
+        return store
+
+    @staticmethod
+    def _extract_trace_path(task: str) -> Path | None:
+        match = re.search(r"(?:^|\s)(?:trace|trace_path)=([^\s]+)", task)
+        if not match:
+            return None
+        raw = match.group(1).strip().strip("\"'")
+        return Path(raw).expanduser() if raw else None
+
+    @staticmethod
+    def _required_str(action: dict[str, Any], key: str) -> str:
+        value = action.get(key)
+        if not isinstance(value, str) or not value.strip():
+            raise ValueError(f"{key} is required")
+        return value.strip()
+
+    @staticmethod
+    def _int_arg(
+        action: dict[str, Any],
+        key: str,
+        default: int,
+        *,
+        minimum: int,
+        maximum: int,
+    ) -> int:
+        value = action.get(key, default)
+        try:
+            parsed = int(value)
+        except Exception:
+            parsed = default
+        return max(minimum, min(maximum, parsed))
+
+
 class DSPyCodingRLMEnvironment(GenericRLMEnvironment):
     """DSPy-focused environment with file edit + tests + DSPy-aware scoring."""
 

{rlm_code-0.1.6 → rlm_code-0.1.7}/rlm_code/rlm/runner.py

@@ -38,6 +38,7 @@ from .environments import (
     GenericRLMEnvironment,
     RLMEnvironment,
     RLMRewardProfile,
+    TraceAnalysisEnvironment,
 )
 from .events import RLMEventBus
 from .frameworks import FrameworkAdapterRegistry, FrameworkEpisodeResult
@@ -279,6 +280,18 @@ class RLMRunner(BenchmarkManagerMixin, ChatSessionMixin, DelegationMixin, Action
                 workdir=self.workdir,
                 reward_profile=self.reward_profile,
             ),
+            "trace_analysis": TraceAnalysisEnvironment(
+                workdir=self.workdir,
+                reward_profile=self.reward_profile,
+            ),
+            "trace-analysis": TraceAnalysisEnvironment(
+                workdir=self.workdir,
+                reward_profile=self.reward_profile,
+            ),
+            "traces": TraceAnalysisEnvironment(
+                workdir=self.workdir,
+                reward_profile=self.reward_profile,
+            ),
             "framework": DSPyCodingRLMEnvironment(
                 workdir=self.workdir,
                 reward_profile=self.reward_profile,

rlm_code-0.1.7/rlm_code/traces/__init__.py

@@ -0,0 +1,6 @@
+"""Trace indexing and query helpers for HALO-style RLM analysis."""
+
+from .index import TraceIndexBuilder
+from .store import TraceStore
+
+__all__ = ["TraceIndexBuilder", "TraceStore"]