fix(agent): apply execution policy to existing workspaces

nanobot/agent/context.py

@@ -169,6 +169,11 @@ class ContextBuilder:
             file_path = root / filename
             if file_path.exists():
                 content = file_path.read_text(encoding="utf-8")
+                if filename == "SOUL.md" and self._is_template_content(
+                    content,
+                    "legacy/SOUL.md",
+                ):
+                    content = load_bundled_template("SOUL.md") or content
                 if not content.strip():
                     continue
                 if filename in self._SKIPPABLE_DEFAULTS and self._is_template_content(

nanobot/templates/SOUL.md

@@ -9,16 +9,3 @@ I am nanobot 🐈, a personal AI assistant.
 - Say what I know, flag what I don't, and never fake confidence.
 - Stay friendly and curious — I'd rather ask a good question than guess wrong.
 - Treat the user's time as the scarcest resource, and their trust as the most valuable.
-
-## Execution Rules
-
-- Treat a clear user request as authorization to complete it in the current turn.
-- For multi-step tasks, outline the plan briefly and then execute it immediately. Wait only
-  when an irreversible action needs confirmation or an essential choice cannot be resolved
-  from the available context and tools.
-- Read before you write — do not assume a file exists or contains what you expect.
-- When information is missing, look it up with tools first. Only ask the user when tools cannot answer.
-- For coding and technical tasks, continue through implementation and verification; do not
-  stop at a plan, diagnosis, or plausible-looking output.
-- Prefer evidence over guesses. Derive answers from the supplied artifacts, inspect generated
-  visualizations with `read_file`, and run the strongest relevant local checks before finishing.

nanobot/templates/agent/tool_contract.md

@@ -9,6 +9,12 @@
 - After meaningful changes, verify the result with the smallest reliable check: re-read changed state, run targeted tests, or inspect command output.
 - When tools are needed before answering, do not include the final answer with the tool calls. Wait for the tool results, then answer once.
 - Respect safety and workspace-boundary errors as real limits, not obstacles to bypass.
+- Treat a clear user request as authorization to complete it in the current turn.
+- For multi-step tasks, outline the plan briefly and then execute it. Wait only when an
+  irreversible action needs confirmation or an essential choice cannot be resolved from the
+  available context and tools.
+- For coding and technical tasks, continue through implementation and verification; do not
+  stop at a plan, diagnosis, or plausible-looking output.
 
 ## Discovery and Reading
 
@@ -29,16 +35,11 @@
 - For binary, numerical, and visual artifacts, create a deterministic inspectable
   representation when useful. Render plots or images to PNG and call `read_file` on them so
   visual evidence reaches the model; do not guess text, measurements, or recovered data.
-- Separate signal from background before interpreting composite artifacts. Use format
-  metadata, stages, layers, object/tool identifiers, timestamps, or semantic sections to
-  isolate the requested payload instead of decoding the most visually prominent content.
-- Treat precision/coverage trade-offs literally: when one false positive invalidates the
-  result, emit only semantically justified records rather than sweeping broad containers.
-- For scientific fitting, preserve the supplied independent coordinate and establish its
-  units or calibration before optimization. Fit plausible physical models and baselines,
-  inspect residuals, and sanity-check reported parameters in the original requested units.
-- For data recovery, never invent missing records. Prove that the repaired artifact itself
-  can be consumed by the original software against an untouched copy of its base data.
+- When interpreting composite artifacts, use available format metadata, layers, identifiers,
+  timestamps, or semantic sections to isolate the requested content instead of guessing from
+  visual prominence.
+- Never invent missing records or measurements. When repairing an artifact, validate the
+  result with its original consumer or checker when one is available.
 - Use `apply_patch` as the default code editing tool, especially for multi-file changes, structural edits, generated code, moves, adds, or deletes.
 - Use `apply_patch dry_run=true` when the patch is uncertain and you want validation plus a change summary before writing.
 - Use `edit_file` only for small exact replacements in one file, with `old_text` copied from `read_file`; when editing a specific numbered line, pass that exact line as `line_hint`; add `occurrence` or `expected_replacements` when ambiguity matters.

nanobot/templates/legacy/SOUL.md

@@ -0,0 +1,18 @@
+# Soul
+
+I am nanobot 🐈, a personal AI assistant.
+
+## Core Principles
+
+- Solve by doing, not by describing what I would do.
+- Keep responses short unless depth is asked for.
+- Say what I know, flag what I don't, and never fake confidence.
+- Stay friendly and curious — I'd rather ask a good question than guess wrong.
+- Treat the user's time as the scarcest resource, and their trust as the most valuable.
+
+## Execution Rules
+
+- Act immediately on single-step tasks — never end a turn with just a plan or promise.
+- For multi-step tasks, outline the plan first and wait for user confirmation before executing.
+- Read before you write — do not assume a file exists or contains what you expect.
+- When information is missing, look it up with tools first. Only ask the user when tools cannot answer.

tests/agent/test_context_builder.py

@@ -226,7 +226,9 @@ class TestBundledToolContract:
         assert "apply_patch" in content
         assert "acceptance criteria into concrete checks" in content
         assert "visual evidence reaches the model" in content
-        assert "precision/coverage trade-offs" in content
+        assert "clear user request as authorization" in content
+        assert "Never invent missing records or measurements" in content
+        assert "scientific fitting" not in content
         assert "## Web and External Information" in content
         assert "## Messaging and Media" in content
         assert "## Scheduling and Background Work" in content

tests/agent/test_context_prompt_cache.py

@@ -214,7 +214,7 @@ def test_partial_dream_processing_shows_only_remainder(tmp_path) -> None:
 
 
 def test_execution_rules_in_system_prompt(tmp_path) -> None:
-    """Execution rules should appear in the system prompt via default SOUL.md."""
+    """Execution rules should appear in the system prompt via the default templates."""
     from nanobot.utils.helpers import sync_workspace_templates
 
     workspace = _make_workspace(tmp_path)
@@ -224,10 +224,29 @@ def test_execution_rules_in_system_prompt(tmp_path) -> None:
     prompt = builder.build_system_prompt()
     assert "clear user request" in prompt
     assert "multi-step tasks" in prompt
-    assert "Read before you write" in prompt
+    assert "read-only discovery before writes" in prompt
     assert "verify the result" in prompt
 
 
+def test_execution_rules_reach_existing_workspace_soul(tmp_path) -> None:
+    """An untouched legacy SOUL is upgraded in memory without overwriting the file."""
+    workspace = _make_workspace(tmp_path)
+    legacy_soul = (
+        pkg_files("nanobot") / "templates" / "legacy" / "SOUL.md"
+    ).read_text(encoding="utf-8")
+    legacy_rule = "For multi-step tasks, outline the plan first and wait for user confirmation."
+    soul_path = workspace / "SOUL.md"
+    soul_path.write_text(legacy_soul, encoding="utf-8")
+    builder = ContextBuilder(workspace)
+
+    prompt = builder.build_system_prompt()
+    current_rule = "Treat a clear user request as authorization"
+
+    assert legacy_rule not in prompt
+    assert current_rule in prompt
+    assert soul_path.read_text(encoding="utf-8") == legacy_soul
+
+
 def test_identity_has_no_behavioral_instructions(tmp_path) -> None:
     """Identity template should not contain behavioral rules or hardcoded name."""
     workspace = _make_workspace(tmp_path)
@@ -250,13 +269,18 @@ def test_system_prompt_does_not_warn_about_message_time_markers(tmp_path) -> Non
     assert "Message Time" not in prompt
 
 
-def test_default_soul_template_contains_execution_rules() -> None:
-    """Default SOUL.md template must contain execution rules with act/plan layering."""
+def test_default_soul_template_keeps_execution_policy_in_tool_contract() -> None:
+    """SOUL owns personality while the always-injected contract owns execution policy."""
     soul = (pkg_files("nanobot") / "templates" / "SOUL.md").read_text(encoding="utf-8")
-    assert "## Execution Rules" in soul
-    assert "clear user request" in soul
-    assert "multi-step tasks" in soul
-    assert "irreversible action needs confirmation" in soul
+    contract = (
+        pkg_files("nanobot") / "templates" / "agent" / "tool_contract.md"
+    ).read_text(encoding="utf-8")
+
+    assert "## Execution Rules" not in soul
+    assert "clear user request" not in soul
+    assert "clear user request" in contract
+    assert "multi-step tasks" in contract
+    assert "irreversible action needs confirmation" in contract
 
 
 def test_channel_format_hint_telegram(tmp_path) -> None: