Tessl • The management layer for AI agent skills

SkillPublic

foresight

Post-implementation lookahead skill. Trigger automatically at the END of any implementation turn — after writing code, modifying files, refactoring, adding features, fixing bugs, or making architectural changes. This skill makes the agent reason one turn ahead: what will break, what edge cases were introduced, what security holes were opened, what patterns are now inconsistent, and what the next human prompt will likely expose as a problem. Use this skill whenever the agent has just shipped a change and is about to respond to the user. It produces a compact, structured FORESIGHT block — never a wall of text. The goal is to surface future breakage *now*, before the next turn, using fewer tokens than the user would spend prompting for it manually. Use this even for small changes — a one-line type change can break ten callers. A one-line config change can expose an entire API.

Quality

86%

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

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.

Highly actionable with a clear, validated workflow and excellent worked examples, but it is verbose for a skill whose output is a 120-word block and keeps all detail inline rather than progressively disclosing it via reference files.

Suggestions

Move the detailed per-category 'How to Reason' breakdowns and the good/bad example sets into a reference file (e.g. references/reasoning-guide.md), keeping SKILL.md a lean overview that points to it.

Trim the Security and Watch-out 'Ask yourself' lists to remove explanations of concepts Claude already knows (injection, CORS, race conditions); keep only the decision prompts.

Eliminate the third restatement of the five categories in 'Self-Check Before Outputting' by referencing the ordered list rather than re-listing each category.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and well-organized, but the ~190-line body over-explains concepts Claude already knows (injection, CORS, race conditions, auth guards) and repeats the five categories across "When to Run", "How to Reason", and "Self-Check"; it could be tightened.	3 / 5
Actionability	Provides a copy-paste-ready FORESIGHT output template, per-category "Ask yourself" prompts, and concrete good/bad worked examples that fully specify what to produce.	5 / 5
Workflow Clarity	Presents an intentional ordered sequence (crash → silent corruption → security → tech debt → next prompt) with an explicit self-check validation checkpoint and a feedback loop ("If you can't answer yes to all five, take 10 more seconds").	5 / 5
Progressive Disclosure	Well-sectioned with clear headers, but with no bundle files the entire ~190-line body is inlined; the detailed per-category reasoning, examples, and calibration could be split into reference files to keep SKILL.md a lean overview.	3 / 5
	Total	16 / 20 Passed

Description

87%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.

A strong description that explicitly states what the skill does and when to run it, with concrete trigger phrasing and a distinctive niche. Minor room to tighten action wording toward more concrete operations and add user-voice synonyms.

Dimension	Reasoning	Score
Specificity	Names several concrete actions ("what will break, what edge cases were introduced, what security holes were opened, what patterns are now inconsistent" and "produces a compact, structured FORESIGHT block"), but they are abstract analytical categories rather than the file-format-specific operations the top anchor exemplifies, so it sits just below comprehensive.	4 / 5
Completeness	Clearly answers both "what" (post-implementation lookahead producing a structured FORESIGHT block) and "when" with concrete trigger phrases ("Trigger automatically at the END of any implementation turn", "Use this skill whenever the agent has just shipped a change").	5 / 5
Trigger Term Quality	Good coverage of natural developer terms ("writing code, modifying files, refactoring, adding features, fixing bugs, ... architectural changes"), but the triggers are framed as the agent's auto-run conditions rather than phrases a user would say, and synonyms are sparse.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (post-implementation lookahead / FORESIGHT block) with distinct triggers and minimal overlap risk against other skills.	5 / 5
	Total	18 / 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: tombrewsviews/foresight-skill
Path: SKILL.md
Revision: 7a95e42

Reviewed: 1 month ago

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.

foresight

Table of Contents

Is this your skill?