菜单
Compute budget management with phase-based termination criteria. Use when checking budget, determining project phase, or making stop decisions.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | budget-tracking |
| description | Compute budget management with phase-based termination criteria. Use when checking budget, determining project phase, or making stop decisions. |
| Phase | Range | Action |
|---|---|---|
| 0 | 0 — minimum | Verify basic operation, fix environment issues |
| 1 | — 50% of target | Aggressive exploration, broad parameter search |
| 2 | 50-80% | Focus on promising approaches, cost-efficiency |
| 3 | 80-100% of target | Best optimization only, start summarizing |
| 4 | target — 90% of deadline | Pre-approve new jobs, prepare final report |
| 5 | 90-100% of deadline | No new jobs, terminate all work within 5 min |
project_usage = current_used - start_used
consumption_% = (project_usage / deadline) × 100
Note: used is cumulative annual value — always diff from project start.
_remote_info/)Details:
references/phase_details.md
determine_phase() is called automatically by summarise() and included in --summary output.DEFAULT_BUDGET_LIMITS keys: Minimum, Expected (target), Deadline.python3 ${CLAUDE_SKILL_DIR}/scripts/budget_tracker.py --help
python3 ${CLAUDE_SKILL_DIR}/scripts/budget_tracker.py /path/to/project --summary
python3 ${CLAUDE_SKILL_DIR}/scripts/budget_tracker.py /path/to/project --report --json
python3 ${CLAUDE_SKILL_DIR}/scripts/budget_tracker.py /path/to/project --jobs --json
python3 ${CLAUDE_SKILL_DIR}/scripts/budget_tracker.py /path/to/project --visualize -o budget.png
ChangeLog.md format spec for VibeCodeHPC. Use when writing or validating ChangeLog entries with compile/job/test/sota fields.
Compiler warning triage for HPC parallelization. Use when deciding whether to proceed with job execution after compile warnings.
Context window usage monitoring and auto-compact recovery. Use when context is filling up or after auto-compact occurs. Optional for strong models.
CLI and dependency installation guidance. Use when setting up VibeCodeHPC on a new environment — local PC, supercomputer, container, etc.
Hardware spec collection and theoretical peak performance calculation for HPC environments. Prerequisite: skills/ssh-management for remote access.
Evolutionary flat-directory exploration strategy for HPC optimization. Use when designing directory hierarchies or planning multi-generation parallelization strategies.