atomise
Atom of Thoughts (AoT) reasoning - decompose complex problems into atomic units with confidence tracking and backtracking. For genuinely complex reasoning, not everyday questions. Triggers on: atomise, complex reasoning, decompose problem, structured thinking, verify hypothesis.
71
64%
Does it follow best practices?
Impact
73%
2.28xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./data/skills-md/0xdarkmatter/claude-mods/atomise/SKILL.mdQuality
Discovery
67%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 is reasonably well-structured with both 'what' and 'when' clauses, including explicit trigger terms and a useful exclusion clause. However, the trigger terms lean technical and may miss common natural language variations users would actually use. The capability description, while naming a methodology, remains somewhat abstract about concrete outputs or actions.
Suggestions
Add more natural trigger terms users would actually say, such as 'break down', 'step by step', 'think through', 'analyze complex problem', or 'systematic reasoning'.
Increase specificity by listing concrete actions or outputs, e.g., 'produces dependency graphs of sub-problems, tracks confidence per sub-conclusion, and backtracks when contradictions are found'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (complex reasoning/problem decomposition) and some actions (decompose into atomic units, confidence tracking, backtracking), but the actions are somewhat abstract rather than concrete. It doesn't list specific output types or detailed steps. | 2 / 3 |
Completeness | Clearly answers both 'what' (decompose complex problems into atomic units with confidence tracking and backtracking) and 'when' (for genuinely complex reasoning, not everyday questions, with explicit trigger terms listed). The anti-trigger ('not everyday questions') adds useful scoping. | 3 / 3 |
Trigger Term Quality | Includes explicit trigger terms ('atomise', 'complex reasoning', 'decompose problem', 'structured thinking', 'verify hypothesis'), but these are somewhat technical/niche. Users are unlikely to naturally say 'atomise' or 'verify hypothesis' — more common terms like 'break down', 'step by step', 'think through', or 'analyze' are missing. | 2 / 3 |
Distinctiveness Conflict Risk | The terms 'structured thinking' and 'complex reasoning' could overlap with other reasoning or problem-solving skills. However, 'atomise' and 'atomic units' provide some distinctiveness. The description could conflict with general chain-of-thought or step-by-step reasoning skills. | 2 / 3 |
Total | 9 / 12 Passed |
Implementation
62%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill provides a well-structured reasoning framework with clear workflow phases, confidence tracking, and backtracking logic. Its main weaknesses are the lack of a concrete worked example showing the full decomposition process in action, and some verbosity in sections that restate or explain concepts Claude would already understand. The workflow clarity is strong with explicit validation checkpoints and error recovery paths.
Suggestions
Add one concrete worked example showing a real problem fully decomposed into atoms with confidence values, verification steps, and the final output table — this would dramatically improve actionability.
Remove the 'Remember' section (restates prior content), trim the 'Honest Caveat' paragraph, and cut the chain-of-thought comparison in the intro to improve conciseness.
Show the actual reasoning trace for at least one of the example invocations (e.g., the Redis vs Memcached one) so Claude can see the expected output format in practice.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient with good use of tables and structured formatting, but includes some unnecessary content like the 'Honest Caveat' explanation about heuristic confidence, the anti-patterns section explaining what not to do, and the 'Remember' summary that largely restates what was already covered. The comparison to chain-of-thought in the intro is also unnecessary context for Claude. | 2 / 3 |
Actionability | The skill provides a clear conceptual framework with structured phases and atom types, but lacks executable code or concrete worked examples. The 'Examples' section only shows invocation commands without showing actual decomposition output. A single worked example showing a real problem decomposed into atoms with confidence tracking would significantly improve actionability. | 2 / 3 |
Workflow Clarity | The multi-step process is clearly sequenced with Phase 0 (Setup) and Phase 1+ (Iterate) stages, explicit validation checkpoints (confidence thresholds, verification adjustments), and a well-defined backtracking protocol with clear triggers (confidence < 0.5) and recovery steps (prune, restore, try alternative). The core loop is concise and unambiguous. | 3 / 3 |
Progressive Disclosure | The content is well-organized with clear sections and headers, but everything is in a single monolithic file. The atom type reference table, mode details, and execution guide could benefit from being split into separate reference files. However, since no bundle files are provided, the inline approach is acceptable though the file is quite long. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- NeverSight/skills_feed
- Commit
2d09b29
- Reviewed
Table of Contents
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.