diff --git a/README.md b/README.md index 14ca5a3..b47b0f7 100644 --- a/README.md +++ b/README.md @@ -35,6 +35,7 @@ python -m esass_prototype.cli --help ``` **Note**: If using `uv` (modern Python package manager): + ```bash pip install uv uv sync @@ -129,30 +130,30 @@ The probe system provides 8 specialized observers (3 core + 5 enhanced): #### Enhanced Meta-Cognitive Probes (NEW in v0.2.0) -4. **ErrorRecoveryProbe** (`esass/probes/error_recovery_probe.py`) +1. **ErrorRecoveryProbe** (`esass/probes/error_recovery_probe.py`) - Detects error → fix → retry patterns - Tracks recovery strategies and success rates - Monitors time-to-recovery metrics -5. **StrategyShiftProbe** (`esass/probes/strategy_shift_probe.py`) +2. **StrategyShiftProbe** (`esass/probes/strategy_shift_probe.py`) - Captures pivot points in problem-solving - Tracks approach changes (analytical → pragmatic) - Measures strategy effectiveness - **Test result**: 1 observation, 4 events generated -6. **CalibrationProbe** (`esass/probes/calibration_probe.py`) +3. **CalibrationProbe** (`esass/probes/calibration_probe.py`) - Compares predictions vs actual outcomes - Tracks confidence calibration over time - Enables self-improvement through feedback - **Test result**: 2 observations, 12 events generated -7. **InsightProbe** (`esass/probes/insight_probe.py`) +4. **InsightProbe** (`esass/probes/insight_probe.py`) - Detects "aha moments" and breakthroughs - Captures high-confidence realizations (>0.7) - Tracks learning and pattern recognition - **Test result**: 1 observation, 4 events generated -8. **ScopeExpansionProbe** (`esass/probes/scope_expansion_probe.py`) +5. **ScopeExpansionProbe** (`esass/probes/scope_expansion_probe.py`) - Identifies complexity surprises - Tracks scope creep and underestimation - Monitors task complexity evolution @@ -178,13 +179,16 @@ Run the integration examples to see the probe system in action: #### Claude Code Example -Run the integration example to see the probe system in action: +Run the integration examples to see the probe system in action: + +#### Claude Code Example ```bash python -m examples.claude_code_integration ``` Expected output: + ```text ====================================================================== ESASS Claude Code Integration Example @@ -221,6 +225,7 @@ Data directory: data_example ``` **What gets captured:** + - 57 total events (45 unique) across 8 event types - Causality chains linking tool calls to outcomes - Semantic tags auto-extracted from parameters @@ -303,6 +308,7 @@ pytest tests/test_probes.py::TestToolCallProbe -v ``` **Test Results Summary**: + ``` ============================= test session starts ============================= platform win32 -- Python 3.9.13, pytest-8.4.2, pluggy-1.6.0 @@ -314,6 +320,7 @@ tests\test_probes.py ........................... [100%] ``` **Detailed test analysis**: See [TEST_RESULTS.md](TEST_RESULTS.md) for: + - Complete test breakdown by component - Event capture examples - Performance metrics