feat(go): add Phase 1 parity sidecar

README.md

@@ -64,6 +64,12 @@ Project docs:
 > detectors, and JSON output. The bash version and the Python version share one
 > red-team harness, so every signature is exercised against both.
 
+Phase 1 now also has an **experimental Go sidecar** under `go-agent/`. It emits
+the same `enodia.event.v1` envelope, captures an injectable `/proc` process
+view, and ports `deleted_exe` as the first parity-checked poll detector. It is
+stdout-only and does not replace, install over, or write state for the Python
+daemon; Python remains authoritative until subsystem parity is complete.
+
 ## Why these detectors
 
 Every detector keys on a behavior that is **cheap to observe** and **expensive
@@ -162,6 +168,14 @@ enodia_sentinel/
     └── monitor.py      runs the probe on a thread, routes events → rules
 ```
 
+The parallel Phase 1 tree is intentionally isolated:
+
+```text
+go-agent/
+├── cmd/enodia-sentinel-go/  experimental JSONL sidecar
+└── internal/                config · model · system · detectors · schema · loop
+```
+
 Three complementary detection paths feed one Alert → snapshot pipeline:
 
 - **Poll** — every few seconds, sweep `/proc`/`ss` (catches anything lingering).
@@ -290,6 +304,8 @@ drill only sets the env var on a throwaway process.
 
 ```bash
 make test          # stdlib unittest suite, no runtime deps
+make test-go       # stdlib-only Go sidecar unit tests
+make parity-go     # shared fixture: Go output vs Python oracle
 ```
 
 Detectors are pure functions over an injectable `SystemState`, so tests build