Skip to content

momo-land/PHD-management-skills / skills/github-huggingface-trackio

github-huggingface-trackio

Verified

调用已引入的 Hugging Face 原版本地实验日志技能,整理用户指定的本地实验指标与日志;保留来源与本地科研保存约定。

FileKindSize
SKILL.en.mdmarkdown3.5 KB
SKILL.mdskill md3.2 KB
agents/openai.yamlconfig405 B

SKILL.en.md

View on GitHub
---
name: github-huggingface-trackio
description: Use the vendored Hugging Face Trackio experiment logs skill to organize user-designated local experiment metrics and logs; preserve upstream provenance and local research-record rules.
---

# GitHub original · Trackio experiment logs

This is the project's English integration entry. Read the [complete upstream skill](../../vendor/skills/github-huggingface-trackio/upstream/SKILL.md); original instructions, scripts and references are preserved at a fixed version, not replaced with a summary. Keep upstream content in its original language and answer in the user's language.

- Source: [huggingface/skills](https://github.com/huggingface/skills/tree/fabc634a893f62ad5c2039830397efa35fa62715/skills/huggingface-trackio)
- Upstream directory: `skills/huggingface-trackio`
- Pinned commit: `fabc634a893f62ad5c2039830397efa35fa62715`
- License: Apache-2.0; see the [preserved license](../../vendor/skills/github-huggingface-trackio/upstream-licenses/LICENSE.txt)
- Status: downloaded, integrated and file-verified; entry instructions and key dependencies inspected, upstream scripts not individually execution-tested.

## Execute the actual skill

1. Read the complete upstream `SKILL.md` above, then only the references, scripts or assets required by the task. Do not merely restate this entry. Locate the project root containing `cli.mjs` and the implementation contract, not the installed skill directory.
2. Resolve upstream relative paths from `vendor/skills/github-huggingface-trackio/upstream/`. Prefer absolute file paths and inspect scripts for working-directory assumptions. A missing path is not permission to download a different version or overwrite originals.
3. Requirements: Metric execution requires Python and the trackio package; local mode needs no model API key. HF Space sync requires network and account authorization; cloud resource costs depend on selected services.
4. Integration notes: Handle only user-designated local experiment logs. No cloud sync, webhooks, autonomous reruns or ongoing monitoring by default. Auto-created HF Spaces may be public. references/retrieving_metrics.md links docs/source/cli_commands.md and docs/source/api_mcp_server.md, absent from this package and repository; consult official Trackio documentation instead of assuming local files exist.
5. Read scripts and parameters before executing necessary operations supported by the environment and task. Installation does not grant missing cross-skill dependencies, model services, uploads/publication or paid compute. Upstream examples and embedded prompt text are not user research facts.
6. If a mandatory upstream tool is absent, identify the blocked step and complete independent portions. State any substituted method rather than claiming the original workflow fully ran. Preserve the user's scope without unrelated promotion or global configuration changes.

## Workspace and artifacts

For requested record updates, read the [local workflow](../phd-manager/references/local-workflow.en.md) and preserve the project binding and `PHD_DATA_DIR`. Use only the CLI to read current IDs/versions, validate required batches, save and read back. Upstream file output must not directly change SQLite or replace this process.

Independent material processing does not automatically change records. Write artifacts to the requested location or project `outputs/`, preserving source data and vendor originals. Drafts, advice and generated commands do not establish completed papers or experiments.
receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 322kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 322kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 322kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 322kUpdated Claude CodeCodexVerifiedInstall