feat: add skills for issue verification and PR review

[generspooler]

Description:
This PR adds two AI-assisted development skills
that enhance the LMCache-Ascend development workflow:

Changes

1. .gitignore: Un-ignore .claude/ directory so that skill files
   can be tracked in version control.

2. .claude/skills/issue-verification-workflow/: An end-to-end workflow
   that automates issue-driven verification and fix pipelines — from
   fetching GitHub issues/PRs, locating root causes, reproducing,
   fixing, to outputting a structured verification report.

3. .claude/skills/pr-review/: A structured PR review workflow that
   performs layered review (conventions, correctness, security,
   architecture, testing, documentation) and generates a review report
   with actionable findings.

[gemini-code-assist]

Code Review

This pull request introduces two automated workflows under .claude/skills/: an issue verification workflow and a PR review workflow, along with their associated scripts, templates, and rubrics. The .gitignore is also updated to track the .claude/ directory. Key feedback on these additions includes: in verify_issue.py, adding import torch_npu to properly initialize the NPU backend, dynamically resolving the repository root instead of hardcoding it, ensuring stream synchronization is called after asynchronous copy operations in benchmarks, and correcting the file paths in the usage docstring; in collect_pr_context.sh, fixing a potential script failure when gh pr list returns "null" for empty PR lists; in the PR review skill documentation, correcting the invalid --request flag to --request-changes for the gh pr review command; and in the issue verification skill documentation, adding a timeout to the health check loop to prevent infinite hangs.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

Hardcoding the absolute path /mnt/sdb/csy/p2p/LMCache-Ascend as DEFAULT_BASE makes the script non-portable and prone to failure on other machines or containers. It is better to dynamically resolve the repository root relative to the script's location.

Suggested change

	DEFAULT_BASE = "/mnt/sdb/csy/p2p/LMCache-Ascend"
	DEFAULT_BASE = str(Path(__file__).resolve().parents[4])

[gemini-code-assist]

Since non_blocking=True is used, the copy operation is asynchronous. Without calling torch.npu.synchronize() after the loop (before taking the end timestamp), the benchmark only measures the CPU launch overhead rather than the actual NPU transfer time, leading to incorrect speedup results.

Suggested change

	torch.npu.synchronize()
	t0 = time.perf_counter()
	for _ in range(n_iters):
	npu_p.copy_(pinned, non_blocking=True)
	t_new = (time.perf_counter() - t0) / n_iters * 1e6
	torch.npu.synchronize()
	t0 = time.perf_counter()
	for _ in range(n_iters):
	npu_p.copy_(pinned, non_blocking=True)
	torch.npu.synchronize()
	t_new = (time.perf_counter() - t0) / n_iters * 1e6

References

1. Do not remove explicit NPU stream synchronization if it is necessary for correctness, even if it limits host-NPU task overlap.

[gemini-code-assist]

If there are no open PRs for the current branch, gh pr list returns an empty array []. The jq expression .[0].number will output the string "null". Since "null" is a non-empty string, [ -n "${PR_NUM}" ] will evaluate to true, causing the script to attempt to run gh pr view "null", which will fail. To fix this, check that PR_NUM is not "null" or use // empty in the jq expression.

Suggested change

	PR_NUM="$(gh pr list --head "${CURRENT_BRANCH}" --state open \
	--json number --jq '.[0].number' 2>/dev/null || true)"
	if [ -n "${PR_NUM}" ]; then
	PR_NUM="$(gh pr list --head "${CURRENT_BRANCH}" --state open \
	--json number --jq '.[0].number // empty' 2>/dev/null || true)"
	if [ -n "${PR_NUM}" ] && [ "${PR_NUM}" != "null" ]; then