task-fix

A repair process for a benchmark task package after its verification checks find problems. It applies targeted fixes, gives failed checks priority over warnings, and tests the package again.

In plain words
What is it for?
Use it to fix problems in the problem description, evaluation, metadata, environment, or other task files, then create a log of the fixes and retest the package.
Why use it?
It turns verification findings into concrete corrections so the task package is more likely to be consistent and runnable.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/frontisai/naturebench/task-fix
Any agent
npx skills add FrontisAI/NatureBench --skill task-fix
Clone the repo
git clone --depth 1 https://github.com/FrontisAI/NatureBench

Made for: Claude Code, Codex.

Per session 49 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,641 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00049 $0.02641
Opus 5 $0.00024 $0.01321
Sonnet 5 $0.00010 $0.00528
Haiku 4.5 $0.00005 $0.00264

Measured 2d ago against content hash 6c3cea56e017, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

task-fix scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

naturegym/.claude/skills/task-fix/SKILL.md · 231 lines

How it starts

The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Task-Fix Skill

Fix issues in a task package identified by task-verify, following task-build rules and conventions.

Input Requirements

Before invoking this skill, provide:

  1. Task Package Path: Directory containing the task package (problem/, evaluation/, environment/, metadata.json)
  2. Paper Folder Path: Directory containing the original paper and prior processing results (needed for context)

The task package directory must contain:

  • task_verify_result.json: Output from task-verify (the issues to fix)

The paper folder must contain:

  • {paper_id}.pdf and {paper_id}.html: Original paper files
  • preprocessed/: paper-preprocess output (text.md, links.json, figures/, tables/)
  • filter_result.json: Combined output from paper-filter + data-check
  • repositories/: Cloned repositories (reference only)

NOTE: The task package path and paper folder path may be the same directory (task-build often builds in-place).

Output

CRITICAL: This skill MUST output both items below, even if no fixes were applied:

  1. Fixed task package: Modified files in the task package directory
  2. Fix log: task_fix_log.txt written to the task package directory

Core Principles

1. Targeted Fixes, Not Rebuilds

This skill makes targeted, minimal fixes to resolve specific verification failures. It does NOT rebuild the task package from scratch — that is task-build's job. Each fix should change only what is necessary to resolve the identified issue.

2. Priority Order

Process issues in this order:

  1. Failed checks — must be resolved (these cause overall_status: "failed")
  2. Warnings — should be resolved where possible (non-critical but improve quality)

3. Fix Dependency Order

Within each priority level, fix in phase order since later fixes may depend on earlier ones:

  • Phase 0 (file structure) → Phase 1 (consistency) → Phase 2 (information firewall) → Phase 3 (benchmark design) → Phase 4 (dynamic testing)

Read the full file on GitHub · 231 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 231 lines · 49 tokens per session scan A 6c3cea56e017

Subscribe to this mod's changes

task-fix is a skill published in the GitHub repository FrontisAI/NatureBench (106 stars, last pushed 2d ago), licensed MIT. It adds 49 tokens to every session and 2,641 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

agent-host-chat-contributions

Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.

microsoft/vscode · 56 tokens

auto-perf-optimize

Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.

microsoft/vscode · 62 tokens