tdg-personal/connections-optimizer
Reorganize the user's X and LinkedIn network with review-first pruning, add/follow recommendations, and channel-specific warm outreach drafted in the user's real voice. Use when the user wants to clean up following lists, grow toward current priorities, or rebalance a social graph around higher-signal relationships.
73
73%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Quality
Discovery
85%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong description that clearly defines a specific niche (social network management on X and LinkedIn) with concrete actions and an explicit 'Use when' clause. Its main weakness is trigger term coverage—it uses some jargon ('higher-signal relationships', 'social graph') that users wouldn't naturally say and misses common synonyms like 'Twitter', 'unfollow', or 'connections'.
Suggestions
Add common user-facing synonyms such as 'Twitter', 'unfollow', 'connections', 'who to follow', and 'networking' to improve trigger term coverage.
Replace jargon like 'higher-signal relationships' and 'social graph' with more natural language users would actually use, such as 'relevant contacts' or 'network'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: review-first pruning, add/follow recommendations, channel-specific warm outreach drafting in the user's voice. These are distinct, actionable capabilities. | 3 / 3 |
Completeness | Clearly answers both what (reorganize X/LinkedIn network with pruning, recommendations, outreach drafting) and when (explicit 'Use when' clause covering cleaning up following lists, growing toward priorities, rebalancing social graph). | 3 / 3 |
Trigger Term Quality | Includes some natural terms like 'X', 'LinkedIn', 'following lists', 'social graph', but misses common user phrasings like 'unfollow', 'Twitter', 'connections', 'networking', 'who to follow'. The phrase 'higher-signal relationships' is jargon users wouldn't naturally say. | 2 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche combining social network management on specific platforms (X and LinkedIn) with pruning, recommendations, and outreach drafting. Unlikely to conflict with other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
50%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-organized framework for network optimization with good safety defaults and a clear review-first philosophy. Its main weaknesses are the lack of executable tool invocation examples (the tools referenced appear custom with no calling conventions shown) and missing validation/error-recovery steps in the workflow. The content is moderately concise but could be tightened by removing signals Claude could infer and consolidating redundant safety mentions.
Suggestions
Add concrete tool invocation examples showing how to call x-api, lead-intelligence, brand-voice, and social-graph-ranker with actual parameters or schemas, so Claude knows exactly how to use them.
Add explicit validation checkpoints in the workflow—e.g., verify graph pull completeness before scoring, validate draft quality against brand-voice output, and define error recovery if API access fails.
Consider moving the scoring model and platform rules into separate referenced files to reduce the main skill's token footprint and improve progressive disclosure.
Tighten the scoring model section—many signals (reciprocity, recent activity, abandoned accounts) are intuitive enough that a brief summary or table would suffice instead of two enumerated lists.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is reasonably well-structured but includes some sections that could be tightened—e.g., the scoring model lists obvious positive/negative signals that Claude could infer, and the platform rules section restates some safety defaults. However, it avoids egregious over-explanation and most content is domain-specific configuration rather than general knowledge. | 2 / 3 |
Actionability | The skill provides a clear workflow sequence, a concrete output format template, and specific tool references, but lacks executable code or commands. The tool names (x-api, lead-intelligence, social-graph-ranker, brand-voice) appear to be custom/hypothetical with no concrete invocation syntax, making it unclear how Claude would actually call them. The guidance is specific in intent but incomplete in execution details. | 2 / 3 |
Workflow Clarity | The 8-step workflow is clearly sequenced and includes a review gate before any apply step, which is good for a destructive operation. However, there are no explicit validation checkpoints within the workflow (e.g., what to do if the graph pull fails, how to verify scoring accuracy, or feedback loops for draft quality). The review pack is the only checkpoint, and error recovery paths are absent. | 2 / 3 |
Progressive Disclosure | The skill references related skills (brand-voice, social-graph-ranker, lead-intelligence, x-api, content-engine) at the end, which is good navigation. However, the main file is fairly long (~150+ lines) with inline content that could potentially be split (e.g., scoring model, platform rules, modes could be separate references). The references are listed but not clearly signaled as links to separate files. | 2 / 3 |
Total | 8 / 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 | |
Reviewed
Table of Contents