Destynova2/cli-code-skills / cli-audit-xray
cli-audit-xray
UnverifiedReveal hidden semantic, mathematical, dataflow, resource-flow, and operational structure in code. Produces optimization candidates with explicit invariants, risks, validation methods, and handoffs. Use when analyzing a function, hot path, AI inference pipeline, compiler/runtime bottleneck, DevSecOps pipeline, LLM gateway, distributed system, or complex business logic for missed optimization opportunities. Also triggers on "semantic xray", "semantic optimization graph", "hidden structure", "optimization cards", "dataflow optimization", "resource flow", "repeated computation", "unnecessary allocation", "cache placement", "CPU/GPU boundary", "prove this optimization", "is this rewrite safe", or "find optimizations humans miss".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.9 KB |
| references/examples.md | markdown | 2.6 KB |
| references/optimization-families.md | markdown | 2.5 KB |
| references/risk-rules.md | markdown | 2.1 KB |
| references/semantic-optimization-graph.md | markdown | 2.9 KB |
references/semantic-optimization-graph.md
View on GitHub# Semantic Optimization Graph
> **When to read:** During `cli-audit-xray` Step 1 when building the semantic map.
The Semantic Optimization Graph (SOG) is a reasoning model, not necessarily a literal file. Build it only as deeply as the scope demands.
## Node Types
| Node | Meaning | Examples |
|---|---|---|
| Operation | A transformation or action | parse, normalize, map, reduce, scan, route, authorize |
| Data | Values moving through the system | request, tensor, token buffer, AST, config, policy result |
| Control flow | Branching, looping, retrying, waiting | if, match, loop, async await, retry, cancellation |
| Resource | Runtime cost carrier | allocation, copy, lock, file, socket, GPU buffer, DB connection |
| Constraint | Fact that limits valid rewrites | order must be stable, tenant isolation, shape equality |
| Invariant | Property expected to always hold | input length = output length, normalized before cache |
| Cost | Runtime or operational burden | O(n^2), extra pass, network hop, kernel launch |
| Risk | What may break | side effects, FP order, exception ordering, auth boundary |
| Candidate | Possible rewrite | fuse maps, cache after canonicalization, preallocate |
## Minimal SOG Card
```yaml
node: op_42
kind: map
source: src/pipeline.rs:118
input: normalized_tokens
output: scored_tokens
properties:
pure: maybe
deterministic: likely
parallelizable: likely
vectorizable: maybe
constraints:
- input_len = output_len
- no_external_state_access_required
cost:
complexity: O(n)
allocates_intermediate: true
memory_passes: 1
risk:
side_effect_risk: medium
floating_point_order_risk: low
behavior_change_risk: medium
optimization_candidates:
- fuse_with_next_map
- remove_intermediate_vec
- parallel_iter
```
## Three Required Views
### Dataflow
Answer:
- Where does data enter?
- How is it transformed?
- Where is it canonicalized?
- Where is it cached?
- Where does it cross trust, process, network, or device boundaries?
- Where is the same value recomputed?
### Control Flow
Answer:
- Which branches change resource usage?
- Which loops contain invariant work?
- Which retries amplify cost?
- Which async waits serialize work that could overlap?
- Which cancellation or exception order is observable?
### Resource Flow
Answer:
- Where are allocations and copies created?
- Where are buffers materialized?
- Where are formats converted?
- Where are locks held?
- Where are network, database, filesystem, or GPU boundaries crossed?
- Where is serialization/deserialization repeated?
## Mapping Rules
- Prefer exact `file:line` evidence over inferred global claims.
- When a node's purity or determinism is unknown, mark it `maybe` and expose the missing invariant.
- Treat security checks, audit logs, metrics, and errors as observable behavior.
- Do not collapse two operations just because names look similar; prove or state the equivalence condition.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 467k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 467k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 467k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 193k | Updated | Claude CodeCodex | Verified | Install |