ScalaSemantic: Skill for Claude Code

.claude/skills/task-splitting-evaluation/SKILL.md

task-splitting-evaluation is a skill for Claude Code from MercurieVV/ScalaSemantic. It costs 73 tokens per session (1,518 once invoked), scanned A, original, MIT.

A process for reviewing GitHub issues before coding begins and turning broad work into smaller, clearly described tasks.

In plain words
What is it for?
Use it to evaluate existing issues, mark simple tasks as ready, split larger tasks into subtasks, and choose a worker for each leaf task.
Why use it?
It prevents agents from starting with vague requirements and avoids processing the same issue more than once.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: reads .claude/ paths.

This is MercurieVV/ScalaSemantic's own configuration. It tells Claude Code how to work on ScalaSemantic itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ScalaSemantic configures →

Reuse

Borrowing it

Nothing to install: this file belongs to MercurieVV/ScalaSemantic. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/MercurieVV/ScalaSemantic/master/.claude/skills/task-splitting-evaluation/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/MercurieVV/ScalaSemantic

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for task-splitting-evaluation

README.md
[![agentmods](https://agentmods.dev/badge/skills/mercurievv/scalasemantic/task-splitting-evaluation/github.svg)](https://agentmods.dev/skills/mercurievv/scalasemantic/task-splitting-evaluation)
Your own site
<a href="https://agentmods.dev/skills/mercurievv/scalasemantic/task-splitting-evaluation"><img src="https://agentmods.dev/badge/skills/mercurievv/scalasemantic/task-splitting-evaluation/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for task-splitting-evaluation

Your own site · 80×15
<a href="https://agentmods.dev/skills/mercurievv/scalasemantic/task-splitting-evaluation"><img src="https://agentmods.dev/badge/skills/mercurievv/scalasemantic/task-splitting-evaluation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 73 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,518 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 1 finding, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • medium Agent Snooping · line 92
    Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
    Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
How audits are shown
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 5.1 $0.00073 $0.01518
Opus 5 $0.00036 $0.00759
Sonnet 5 $0.00015 $0.00304
Haiku 4.5 $0.00007 $0.00152

Measured 9d ago against content hash 111ef920d3cf, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

task-splitting-evaluation 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 9d 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.

.claude/skills/task-splitting-evaluation/SKILL.md · 144 lines

How it starts

The opening of the file, as written. The whole thing — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Task Splitting And Evaluation

Use this skill before the implementation-oriented orchestrate skill. This flow does not write code, create branches, or merge PRs. It builds a bounded GitHub issue tree so every executable leaf has a clear task description and selected worker engine/model.

Agents

  • .claude/agents/task-tree-triage.md - cheap Haiku classifier for one issue. It decides whether to skip, make the issue an executor-ready leaf, or send it to the smart planner for subtask expansion.
  • .claude/agents/task-plan-architect.md - strongest-model expansion planner. It inspects one broad issue, creates bounded GitHub subtasks, selects a preferred engine/model for each subtask, and comments the parent with the generated task tree.

Processing Marker

Every processed issue must get exactly one durable marker comment unless it is skipped because an equivalent marker or executor-ready comment already exists:

### Agentic task-tree marker

Status: leaf-ready|expanded|skipped
Processed-by: task-splitting-evaluation-v1
Tree depth: <n>

This marker is the loop guard. Never process an issue twice in the same run, and never process an issue that already has this marker in its body or comments.

Outcomes

For each unprocessed issue, choose exactly one outcome:

  1. leaf-ready

    • Use when the task can be implemented directly by one worker.
    • Add a task comment with a detailed implementation-ready description if the issue does not already have one.
    • Select the preferred executor engine/model from .claude/orchestrate-routing.md.
    • The issue becomes a tree leaf.
  2. expanded

    • Use when the task is broad, vague, risky, or naturally decomposes into dependent steps.
    • Run .claude/agents/task-plan-architect.md with the smartest available LLM/model. Do not cost-optimize the planning model for this step.
    • The architect creates one GitHub issue per subtask and links it to the parent as a native GitHub sub-issue with GraphQL addSubIssue.
    • Comment the parent with child issue links, dependency order, and per-child executor/model choices.
    • Newly created subtasks enter the queue and are processed by the same rules.

Read the full file on GitHub · 144 lines

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. 9d ago First seen · 144 lines · 73 tokens per session scan A 111ef920d3cf

Subscribe to this mod's changes

task-splitting-evaluation is a skill published in the GitHub repository MercurieVV/ScalaSemantic (21 stars, last pushed yesterday), licensed MIT. It adds 73 tokens to every session and 1,518 once invoked, about $0.0004 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

roam

Codebase comprehension via roam-code CLI. Use when exploring codebases, planning modifications, debugging failures, assessing PR risk, or checking architecture health. Triggers on: understanding project structure, pre-change safety checks, finding symbols/files, blast radius analysis, affected tests, health scoring…

Cranot/roam-code · 86 tokens

mcpls

Install, configure, and run the mcpls CLI — a Rust binary bridging MCP to LSP that gives an agent compiler-grade code intelligence (hover, definitions, references, diagnostics, rename). Use when setting up or registering mcpls as an MCP server, writing or debugging an mcpls.toml, choosing CLI flags or MCPLS…

bug-ops/mcpls · 89 tokens

mastermind-task-planning

Choose the lightest Mastermind workflow that fits the risk, then create an evidence-grounded verified or strict task contract for delegated implementation. Direct work deliberately uses no task spec.

xcrft/mastermind · 41 tokens

mastermind-comment-audit

Read-only review of a non-empty comment delta. Flags narration with evidence, names what it deliberately kept, and reports removed rationale. Use when the user asks to check comments or a finished diff's pre-gate found added, modified, or deleted comments; skip automatic routing when no comment changed.

xcrft/mastermind · 65 tokens

mastermind-test-audit

Read-only review of whether the tests in a finished change actually prove its behaviour — changed code no test reaches, a test exercising a different path than the one that changed, an assertion edited to match the new output, and a suite that ran nothing. Use after implementation, and whenever a change claims to be…

xcrft/mastermind · 69 tokens

mastermind-architecture-review

Review a proposed or existing architecture against the real runtime path, source-of-truth ownership, retry and idempotency behavior, and backward-compatibility constraints. Use for service boundaries, async workflows, persistence changes, external integrations, migrations, public APIs, events, schemas, or designs…

xcrft/mastermind · 76 tokens