feat(adapter/amazon-q): Amazon Q Developer CLI skill runtime

**Goal:** Bring Amazon Q Developer CLI (`q`) to feature parity with Claude Code as a skill runtime. Grounds: [RFC-AI-0004 Principle 3 — Vendor neutrality](../tree/main/docs/rfcs/RFC-AI-0004.md).

**Why Amazon Q:** Default agentic CLI in many AWS-shop environments — corporate adopters, federal / regulated environments where Amazon Q is the only sanctioned LLM tool. Several ASF projects with AWS-aligned committer bases have asked about Q compatibility. Coverage matters less for individual maintainers, more for *organisational* adoption.

**What "parity" means:**

- Skills under `.claude/skills//SKILL.md` are invokable from `q` (Q Developer CLI)
- The `tools/*` bridges are reachable — `q` supports custom tools via its agent configuration
- Sandbox parity: `q` has its own command-permission model (allow-lists, "trust" prompts) — document the mapping
- HITL primitives: `q` supports interactive confirmation for tool use — verify and document

**Suggested approach:**

- Map `SKILL.md` → Q's agent configuration format (`~/.aws/amazonq/agents/.json` + system prompts)
- Document skill-invocation pattern
- Map sandbox parity — `q` allowlists are the native fit for the framework's command-permission discipline

**Reference:**

- RFC-AI-0004: [`docs/rfcs/RFC-AI-0004.md`](../tree/main/docs/rfcs/RFC-AI-0004.md)
- Amazon Q Developer CLI: https://docs.aws.amazon.com/amazonq/latest/qdeveloper-ug/command-line.html
- Q agents: https://docs.aws.amazon.com/amazonq/latest/qdeveloper-ug/cli-agents.html
- Existing skill shape: [`.claude/skills/`](../tree/main/.claude/skills/)

Read docs/rfcs/RFC-AI-0004.md and inspect the existing .claude/skills/ shape first. Consult the Amazon Q CLI and Q agents references, then verify how skills, tools/* bridges, allow-lists, trust prompts, and interactive confirmations map to Q. Done means the mapping and invocation pattern work and the sandbox and HITL behavior are documented.

Written by the indexing model from the issue text.