feat: bundled Go service example and real Go toolchain vertical slice

[howlcipher]

Summary

Implements the roadmap item BUNDLED GO EXAMPLE PROJECT + REAL GO TOOLCHAIN VERTICAL SLICE for DevOps-Learn-by-Doing.

Key Deliverables

1. Bundled Go Example Service (projects/go_service/):
   • Runnable Go HTTP service with /health and /info endpoints using Go standard library (net/http, log/slog).
   • Non-root, multi-stage Dockerfile (golang:1.22-alpine builder + alpine:3.20 runtime).
   • Package-level unit tests (health_test.go, info_test.go) asserting status codes, payload shapes, environment overrides, and secret-safety guarantees.
2. Go Tool Implementation (src/devops_learn/tools/go_tool.py):
   • Concrete RealGoTool and SimulatedGoTool extending Tool with allow-listed operations: run_tests, run_vet, run_build, run_fmt_check, verify_modules, version.
   • Execution protected via _subprocess_safety.run_safely (argument arrays, timeouts, output sanitization, non-zero return code detection, shell=False).
3. Language-Agnostic Project Analysis (src/devops_learn/analysis/project_analyzer.py):
   • Generalized _detect_language to detect LanguageKind.GO via go.mod and *.go sources without folder/path special-casing.
   • Generalized secret indicator and environment variable regexes for Go patterns (os.Getenv, os.LookupEnv).
   • Generalized observability detection for Go logging libraries (slog, log, zap, logrus).
4. Stack-Aware Local Workflow Execution (src/devops_learn/workflows/local_flow.py):
   • Fail-closed execution sequence for Go services: gofmt format check -> go vet static analysis -> go test unit tests -> go build binary build -> docker build multi-stage container -> docker run -> endpoint health check -> docker logs -> docker stop.
   • Contextual learning moments teaching Go-specific concepts (static binary vs interpreted runtime, multi-stage Docker builds, static analysis vs unit testing).
5. Planning & Deploy Integration (planning/service.py, workflows/deploy_flow.py, bootstrap.py, cli/main.py):
   • PlanningService and DeployFlow validation steps dynamically generate and validate language-specific tool steps based on assessment.language.
6. Comprehensive Verification & Falsification Suites:
   • Added tests/tools/test_go_tool.py, tests/troubleshooting/test_go_falsification.py, and tests/workflows/test_local_flow.py.
   • Explicitly tested that Go test failures, go vet failures, compilation errors, malformed modules, and healthcheck failures fail closed.
   • Updated GitHub CI workflow with Go 1.22 toolchain setup, formatting, vet, test, build, and container build steps.

Evidence of Verification

• ai verify: PASSED (flake8 exit 0, pytest exit 0: 199 passed, 1 skipped)
• flake8: Clean (exit 0)
• mypy + demo app mypy app --strict: Clean (exit 0, 147 + 4 source files)
• Live Go toolchain (gofmt, go mod verify, go vet, go test, go build): All passed on projects/go_service/