AGENT SKILL

agents-and-models

Configure and inspect tree-of-thoughts TotAgent model adapters, Thought output contracts, and safe model preflight checks.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
agents-and-models
license
Apache 2.0
metadata
{ "disco-role": "operating" }
description
Configure and inspect tree-of-thoughts TotAgent model adapters, Thought output contracts, and safe model preflight checks.
disable-model-invocation
true

02 资源文件 3

03 SKILL.md 正文

agents-and-models

Use this sub-skill when a task needs to configure, inspect, or preflight the tree-of-thoughts model-facing agent layer before using search workflows. It covers TotAgent, the Thought schema, OpenAI caller configuration, injected custom or fake models, model-output validation, and failure diagnosis.

Route here for

  • Verifying that TotAgent imports and that the installed dependency set is compatible.
  • Choosing between the default OpenAI function caller and use_openai_caller=False with a custom or fake model runner.
  • Validating that a model returns dict-like text with thought and numeric evaluation keys.
  • Building deterministic fake agents for safe downstream DFS/BFS smoke checks without external services.
  • Diagnosing import, API-key, parser, constructor side-effect, or no-PyTorch warnings around model setup.

Do not handle here

  • DFS/BFS traversal parameters, pruning behavior, breadth limits, or result interpretation. Route those to the search-workflows sub-skill.
  • Stale maintainer scripts, packaging release automation, or project-wide cleanup. Route those to the root troubleshooting or maintainer references.

Runtime references

  1. Read references/api-reference.md for verified signatures, install/import expectations, environment setup, model-output contracts, and safe fake/custom model examples.
  2. Read references/troubleshooting.md for concrete symptoms, causes, commands, and expected recovery signals.
  3. Use scripts/check_model_contract.py to validate sample model output before passing it into TotAgent.run or DFS/BFS wrappers.

Operating guardrails

  • Treat model output as trusted-only or tightly constrained: this package converts text to a dict with Python eval.
  • Prefer deterministic fake/custom runners for CI, smoke tests, and offline checks; the default TotAgent(use_openai_caller=True) requires OPENAI_API_KEY and external OpenAI-compatible service access.
  • Keep API keys out of prompts, logs, sample outputs, and generated files.
  • Validate the model contract before debugging DFS/BFS behavior; traversal code assumes every generated thought is a dict with exact thought and evaluation keys.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351