feat: add QClaw agent support

[arbing]

• Register QClaw as a supported OpenClaw wrapper with its own agent type, configuration keys, default directory, and frontend metadata.
• Parse QClaw JSONL transcripts from qclaw.qq.com, including messages, thinking blocks, tool calls, tool results, and token usage.
• Discover active and archived QClaw session files under .qclaw/agents while avoiding duplicate archived transcripts.
• Wire QClaw path classification into incremental sync and cover active/archive classification behavior with tests.
• Document QCLAW_DIR in CLI help so users can point agentsview at custom QClaw data directories.

[roborev-ci]

roborev: Combined Review (8e037ff)

PR has one medium issue to address before merge.

Medium

• scripts/e2e-server.sh:57 - The e2e server wrapper does not isolate QClaw discovery. QClaw now adds a default ~/.qclaw/agents discovery root, but the script only overrides the older agent env vars, so developers with real QClaw sessions may have those sessions discovered during e2e runs. This breaks the script’s isolation guarantee and can make tests nondeterministic.

  Fix: Add QCLAW_DIR="$EMPTY_DIR" \ alongside OPENCLAW_DIR in scripts/e2e-server.sh.

---

Synthesized from 3 reviews (agents: codex, gemini | types: default, security)

[roborev-ci]

roborev: Combined Review (ec0272d)

Summary verdict: One medium issue remains; no high or critical findings were reported.

Medium

• internal/parser/qclaw.go:175 - QClaw toolResult messages compute ContentLength but do not populate ParsedToolResult.ContentRaw. As a result, pairToolResults cannot decode and store/display result content for matching tool calls, so QClaw tool output is silently dropped even for unblocked categories.
  • Fix: Set ContentRaw: content.Raw on ParsedToolResult and add a parser or sync test verifying a QClaw tool result appears on the paired tool call.

---

Synthesized from 3 reviews (agents: codex, gemini | types: default, security)

[roborev-ci]

roborev: Combined Review (ff7f420)

No Medium, High, or Critical findings were reported.

All available review outputs agree the code is clean.

---

Synthesized from 3 reviews (agents: codex, gemini | types: default, security)

[roborev-ci]

roborev: Combined Review (ff7f420)

Clean: all completed reviews report no Medium, High, or Critical findings.

No issues found.

---

Synthesized from 3 reviews (agents: codex, gemini | types: default, security)

[roborev-ci]

roborev: Combined Review (ff7f420)

Verdict: No Medium, High, or Critical issues were reported.

All populated review outputs found the code clean.

---

Synthesized from 3 reviews (agents: codex, gemini | types: default, security)