plugins/beaver/skills/beaver-report/SKILL.md
Generate a project health report with milestone progress, stale/overdue detection, blocking chains, and risk analysis. Trigger when the user asks about project status, progress, health, or risks.
npx skillsauth add primatrix/skills beaver-reportInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Generate a comprehensive project health report covering milestone progress, health indicators, risk analysis, and sub-task rollup.
References beaver-engine for: label taxonomy (Section 1), label ops (Section 4), config reading (Section 5).
gh auth status must succeedRead beaver-config per engine Section 5. Identify issue repo and observed repos.
gh api repos/{org}/{issueRepo}/milestones --jq '[.[] | select(.state=="open") | select(.due_on != null)] | sort_by(.due_on) | .[0] // empty | {number, title, due_on, open_issues, closed_issues}'
gh api repos/{org}/{issueRepo}/issues?labels=Control-By-Beaver\&state=open\&per_page=100 \
--jq '.[] | {number, title, labels: [.labels[].name], assignees: [.assignees[].login], updated_at, created_at, milestone: (.milestone.title // null)}'
Parse each issue's labels per engine Section 4 and check:
Issues where status/in-progress or status/review-needed and updated_at > 3 days ago.
For each: add beaver/stale label if not already present.
Issues with milestone whose due_on has passed and status is not status/done.
For each: add beaver/overdue label if not already present.
Skip issues with beaver/wontfix label.
Scan issue bodies for Depends on #{N} patterns. Check if referenced issues have status/blocked. Mark downstream issues with beaver/upstream-blocked.
Issues without type/ or size/ label (excluding those in status/triage).
For issues with size/L, fetch sub-issues:
gh api repos/{org}/{issueRepo}/issues/{number}/sub_issues \
-H "X-GitHub-Api-Version: 2026-03-10" \
--jq '.[] | {number, title, labels: [.labels[].name]}'
Calculate: total sub-tasks, completed (has status/done), percentage.
# Beaver Project Report
**Generated:** {date} | **Milestone:** {title} (due {due_on})
## Milestone Progress
| Metric | Value |
|--------|-------|
| Total Issues | {open + closed} |
| Closed | {closed} ({pct}%) |
| size/L | {count} ({completed}/{total}) |
| size/S | {count} ({completed}/{total}) |
## Health Indicators
### Overdue ({count})
| # | Title | Assignee | Days Overdue |
|---|-------|----------|-------------|
### Stale ({count})
| # | Title | Status | Days Since Update |
|---|-------|--------|------------------|
### Blocked ({count})
| # | Title | Blocked By |
|---|-------|-----------|
### Missing Context ({count})
| # | Title | Missing |
|---|-------|---------|
## size/L Progress Rollup
| # | Title | Sub-tasks | Completed | Progress |
|---|-------|-----------|-----------|----------|
## Risk Summary
{LLM analysis of the above data: top 3 risks with recommended actions}
Ask user: "Publish this report as an Issue comment? (provide issue number, or skip)"
If yes, write the report to a temporary file first, then post it:
REPORT_FILE=$(mktemp)
cat > "$REPORT_FILE" << 'BEAVEREOF'
{rendered_report}
BEAVEREOF
gh api repos/{org}/{issueRepo}/issues/{target_number}/comments --method POST \
--raw-field body=@"$REPORT_FILE"
rm "$REPORT_FILE"
beaver/wontfix issues for stale/overdue detectiondevelopment
Use when analyzing TPU pretraining HBM occupancy from a profile directory — locates the static HBM peak (the same number TensorBoard's Memory Viewer shows), enumerates every buffer alive at the peak schedule moment with size / HLO instruction / opcode / op_name, and rolls the alive set up by opcode and op_name. Reads compile-time `*.hlo_proto.pb` (BufferAssignmentProto) as the primary source; runtime `*.xplane.pb` allocator events are a secondary, often-truncated signal.
testing
Use when analyzing TPU pretraining compute efficiency from xplane.pb — produces source-line-aggregated HLO duration tables, layer-scoped breakdowns, non-compute (padding/cast/copy) audits, and v7x roofline shortfall vs theoretical peak. Reads schema documented by profile-anatomy.
tools
--- name: comm-analysis description: Use when analyzing communication on a TPU pretraining profile — extracts every comm primitive (async + sync, TC + SparseCore), attributes axes via HLO replica_groups, computes per-row NCCL bus BW vs per-axis peak ICI BW (peak_link × k_torus_dims × directions_per_dim; TPUv7x: 200 GB/s bidir per link on a 3D torus; util% requires `--mesh-spec` with topology), and reports per-step compute/comm overlap. Builds on profile-anatomy. --- # Communication Analysis **
documentation
Use when reading TPU pretraining profiles (xplane.pb, trace.json.gz) — describes the on-disk layout, the XSpace/XPlane/XLine/XEvent/XStat hierarchy, and provides reference scripts that future tpu-perf skills can read as schema documentation.