handle-large-tasks

Use this skill to split large plans into smaller chunks. This skill manages your context window for large tasks. Use it when a task will take a long time and cause context issues.

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.agency/plugins/nori/skills/handle-large-tasks/SKILL.md

Quality

Content

77%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body delivers a genuinely actionable, well-sequenced subagent delegation workflow with strong validation and recovery loops, in a compact form. Its main weakness is redundancy: the Guidelines section partially re-explains the required steps and adds context Claude already has.

Suggestions

Trim or merge the Guidelines section with the required block to remove restated steps (tests, review, final ownership) and known-context filler like 'Each subagent has its own limited context window'.

Add one concrete example of a subagent instruction or test so 'write a test for each subagent' is unambiguous in practice.

Specify how to verify that all code 'fits together coherently' (e.g., a named integration check or run-the-full-test-suite step).

Dimension	Reasoning	Score
Conciseness	The body is short, but the Guidelines section restates the required steps (write tests, review output, own the final result) and explains concepts Claude already knows ('Each subagent has its own limited context window', 'Subagents perform best when they have clear implementation guidelines'). Fits 'mostly efficient but includes some unnecessary explanation or could be tightened'; not 4 because the redundancy between the two sections is noticeable rather than minor.	3 / 5
Actionability	Concrete, executable guidance is given: named tools (TodoWrite, Task tool), a test-driven delegation pattern (write a test, have the subagent make it pass), and explicit iteration and restart handling. Minor gaps remain, e.g., what a subagent test should look like and how to judge that code 'fits together coherently'; not 5 for those unspecified details, not 3 because the instructions are directly executable rather than merely high-level hints.	4 / 5
Workflow Clarity	The sequence is clear (announce, plan, test, start, review, evaluate, integrate) with explicit validation checkpoints (tests must pass) and feedback loops for error recovery ('Iterate until tests pass AND the code fits', restart handling that passes in the previous plan, and a final all-tests-pass integration check). Matches the top anchor.	5 / 5
Progressive Disclosure	The skill is under 50 lines with no need for external references (no references/, scripts/, or assets/ bundles exist), and the content is organized into a required-steps block and a Guidelines section. Per the rubric's simple-skill guideline, this earns the top score with just well-organized sections.	5 / 5
	Total	17 / 20 Passed

Description

53%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description answers both what and when with a clear trigger clause, but it leans on generic phrasing ('large tasks', 'long time') and only one concrete capability. Rewriting the capability statement in third person with additional concrete actions and richer trigger variations would materially improve it.

Suggestions

State capabilities in third person with more concrete actions, e.g., 'Splits large tasks into subagent-sized chunks with tests for each subagent. Manages context by delegating implementation to subagents.'

Broaden trigger terms with natural user phrasings such as 'big task', 'split up the work', 'delegate to subagents', or 'running out of context'.

Make the 'when' clause more concrete by naming the situations that call for this skill (e.g., multi-file implementations, tasks too long for one context window).

Dimension	Reasoning	Score
Specificity	Only one concrete action is stated ('split large plans into smaller chunks'); 'manages your context window for large tasks' is generic. This matches the 1-2-concrete-actions anchor (3), but the capability is stated in second-person imperative form ('Use this skill to split...'), which the rubric penalizes by reducing the specificity score by 1.	2 / 5
Completeness	Both parts are explicitly answered: what ('split large plans into smaller chunks', 'manages your context window') and when ('Use it when a task will take a long time and cause context issues'). Not 5 because the 'when' clause is fairly general and could name more concrete trigger situations; not 3 because an explicit trigger clause is present.	4 / 5
Trigger Term Quality	Relevant keywords are present ('large plans', 'smaller chunks', 'context window', 'large tasks', 'context issues') but common natural variations are missing (e.g., 'big task', 'split up the work', 'delegate', 'subagents', 'running out of context'). Fits the 'some relevant keywords but missing common variations or synonyms' anchor; not 4 because users would often phrase the need differently.	3 / 5
Distinctiveness Conflict Risk	The context-window-management-via-chunking niche is identifiable, but triggers like 'a task will take a long time' are generic and would overlap with general planning, decomposition, or task-breakdown skills. Fits 'somewhat specific but could still overlap with similar skills'; not 4 because the overlap with decomposition/planning skills is more than minor.	3 / 5
	Total	12 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: microsoft/FluidFramework
Path: .agency/plugins/nori/skills/handle-large-tasks/SKILL.md
Commit: e41dc04

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.