| name | managing-pants-build |
| description | Use when working with Pants Build — pants build system management. Covers
BUILD file analysis, target querying, dependency inference, remote execution,
cache configuration, and build performance. Use when managing Pants
workspaces, debugging dependency resolution, analyzing target graphs, or
optimizing build performance.
|
| connection_type | pants |
| preload | false |
Pants Build System Management Skill
Manage and analyze Pants build targets, dependency inference, and execution.
MANDATORY: Discovery-First Pattern
Always check current Pants configuration and target structure before modifying builds.
Phase 1: Discovery
#!/bin/bash
echo "=== Pants Configuration ==="
cat pants.toml 2>/dev/null | head -25
echo ""
echo "=== Pants Version ==="
pants --version 2>/dev/null
echo ""
echo "=== BUILD Files ==="
find . -name "BUILD" -not -path "*/node_modules/*" 2>/dev/null | head -15
echo ""
echo "=== Enabled Backends ==="
grep -A10 '\[GLOBAL\]' pants.toml 2>/dev/null | grep backend | head -10
Phase 2: Analysis
#!/bin/bash
echo "=== All Targets ==="
pants list :: 2>/dev/null | head -20
echo ""
echo "=== Target Types ==="
pants list --output=json :: 2>/dev/null | jq -r '.[].target_type' 2>/dev/null | sort | uniq -c | sort -rn | head -10
echo ""
echo "=== Dependency Graph ==="
pants dependencies --transitive :: 2>/dev/null | wc -l | xargs -I{} echo "Total dependency edges: {}"
Output Rules
- TOKEN EFFICIENCY: Target <=50 lines per output
- Summarize targets by type
- Report dependency graph density concisely
- List backends and plugins enabled
Common Operations
Target Dependencies
#!/bin/bash
TARGET="${1:?Target address required}"
echo "=== Dependencies for $TARGET ==="
pants dependencies "$TARGET" 2>/dev/null | head -15
echo ""
echo "=== Dependees ==="
pants dependees "$TARGET" 2>/dev/null | head -10
Build and Test
#!/bin/bash
echo "=== Changed Targets ==="
pants --changed-since=main list 2>/dev/null | head -10
echo ""
echo "=== Test Changed ==="
pants --changed-since=main --changed-dependees=transitive test 2>&1 | tail -15
Safety Rules
- Use
--changed-since to limit builds to affected targets in CI
- Never disable dependency inference without understanding manual dependency cost
- Test pants.toml changes locally before committing
- Review lockfile updates generated by Pants resolve commands
Output Format
Present results as a structured report:
Managing Pants Build Report
═══════════════════════════
Resources discovered: [count]
Resource Status Key Metric Issues
──────────────────────────────────────────────
[name] [ok/warn] [value] [findings]
Summary: [total] resources | [ok] healthy | [warn] warnings | [crit] critical
Action Items: [list of prioritized findings]
Target ≤50 lines of output. Use tables for multi-resource comparisons.
Anti-Hallucination Rules
- NEVER assume resource names — always discover via CLI/API in Phase 1 before referencing in Phase 2.
- NEVER fabricate metric names or dimensions — verify against the service documentation or
--help output.
- NEVER mix CLI commands between service versions — confirm which version/API you are targeting.
- ALWAYS use the discovery → verify → analyze chain — every resource referenced must have been discovered first.
- ALWAYS handle empty results gracefully — an empty response is valid data, not an error to retry.
Counter-Rationalizations
| Shortcut | Counter | Why |
|---|
| "I'll skip discovery and check known resources" | Always run Phase 1 discovery first | Resource names change, new resources appear — assumed names cause errors |
| "The user only asked for a quick check" | Follow the full discovery → analysis flow | Quick checks miss critical issues; structured analysis catches silent failures |
| "Default configuration is probably fine" | Audit configuration explicitly | Defaults often leave logging, security, and optimization features disabled |
| "Metrics aren't needed for this" | Always check relevant metrics when available | API/CLI responses show current state; metrics reveal trends and intermittent issues |
| "I don't have access to that" | Try the command and report the actual error | Assumed permission failures prevent useful investigation; actual errors are informative |
