Add agentic-coding scaffolding (AGENTS.md, CLAUDE.md, session hook)

[grzanka]

Make the repo first-class for AI coding agents (Claude Code, opencode,
Qwen Code, Codex) so they share one project brief and a reproducible
build/test/lint loop:

• AGENTS.md: single source of truth — build/test/lint commands, repo map,
  API model, enforced conventions, generated-data and thread-safety
  gotchas, definition of done, and how to point opencode at a self-hosted
  Qwen (OpenAI-compatible) endpoint.
• CLAUDE.md: imports AGENTS.md so Claude-specific tooling reads the same brief.
• .claude/hooks/session-start.sh: SessionStart hook for Claude Code on the
  web — configures + builds the debug preset, installs the editable Python
  binding with dev deps, and exports LIBDEDX_SO so ctest/clang-tidy/pytest
  work in-session. Idempotent; remote-only.
• .claude/settings.json: registers the hook and pre-allows the safe,
  repeated dev commands (cmake/ctest/ruff/pytest/clang-*/valgrind/git status).

https://claude.ai/code/session_01PKNN8m7LaiTseSm1RYkudy

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 72.73%. Comparing base (36990c2) to head (4812c28).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #137   +/-   ##
=======================================
  Coverage   72.73%   72.73%           
=======================================
  Files          12       12           
  Lines        1643     1643           
  Branches      300      300           
=======================================
  Hits         1195     1195           
  Misses        448      448           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull request overview

Adds AI-agent onboarding/scaffolding so multiple agent tools can share a single, consistent project brief and an automated “inner loop” setup for Claude Code web sessions.

Changes:

• Add AGENTS.md as a shared, tool-agnostic project brief (build/test/lint loop, repo map, API model, conventions, gotchas).
• Add CLAUDE.md that imports the shared brief for Claude Code.
• Add Claude Code web SessionStart hook + settings to auto-configure/build debug preset, install Python dev deps, and set LIBDEDX_SO.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
CLAUDE.md	Points Claude Code to the shared agent brief (AGENTS.md).
AGENTS.md	Centralized agent-facing documentation for workflows, structure, and conventions.
.claude/settings.json	Registers the session-start hook and declares an allowlist for common dev commands.
.claude/hooks/session-start.sh	Automates configure/build + editable Python install + LIBDEDX_SO export in Claude Code web sessions.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.