Thomas Hoogteijling
Community@blauwtje · Rotterdam
HBO Informatica student building practical software with .NET, TypeScript and AI.
Agent Skills by Thomas Hoogteijling
Showing 22 vetted skills indexed across 1 GitHub repositories.
research
Verifies pinned external library behavior against first-party documentation before code decisions.
settings
Configure exo plugin settings across project, local, and global layers via scripts.
swarm
Fans one job out to parallel delegate workers for coverage, races, gauntlets, or judged arenas.
savings
Relay the read guard's savings report of tokens kept out of context.
technical-writing
Writes and edits documentation prose using plain-language rules and an AI-tics checklist.
memory
Records user corrections about a repository into a two-session-attested project memory file.
blast-radius
Analyzes what a code diff breaks outside its changed lines before merging.
handoff
Writes a session state note to a git-scoped path for resuming work after a context clear.
implementing
Executes plan files task by task through delegated build contexts with per-task commits.
deepen
Audits module architecture and ranks refactoring candidates by evidence read from scoped code.
issuing
Creates GitHub issues written as specs using the repository's own labels, types, and project fields.
planning
Converts a development request into an executable, dependency-ordered implementation plan artifact.
shipping
Push branches, open pull requests, and merge them through GitHub API gates.
investigation
Explains repository code behavior and rationale using session-read evidence and git history.
prototyping
Build throwaway code experiments to answer open logic and data-flow questions.
hillclimb
Optimizes one measured metric toward a target through an unattended keep-or-revert loop.
designing
Designs and builds distinctive visual surfaces through a five-phase loop with rendered critique and QA.
skills-tool
Create, edit, and pressure-test Claude Code skills and agents against model behavior.
implementing-batch
Implements decided code changes across multiple files with ordered edits, proof, and review.
refactoring
Restructures internal APIs and code while pinning behavior with before-and-after output comparisons.
shaping
Interviews open product decisions and writes a buildable brief before implementation.
debug
Diagnose bugs by reproducing symptoms and proving the causal mechanism before fixing.
Frequently Asked Questions About Thomas Hoogteijling
FAQPage SchemaWhat tasks can I automate using Thomas Hoogteijling's exo skills?▼
The exo skills cover the full change lifecycle: planning multi-step edits, implementing decided changes test-first, debugging reported failures, running behavior-preserving refactors, checking blast radius before merges, filing GitHub issues, and pushing or merging pull requests with fixed checks and addressed review comments.
Who are these exo skills designed for?▼
Software engineers working in git repositories who want disciplined, evidence-based sessions. The skills suit developers doing TDD, refactors, dependency bumps, and PR reviews, plus maintainers who need architecture audits (deepen), parallel worker fan-out (swarm), or unattended metric optimization loops (hillclimb).
How do the exo skills run in practice?▼
Each skill is invoked by name with an argument hint, such as a plan path, symptom, or diff. Several use scoped Bash permissions on node scripts (settings.mjs, savings.mjs, memory.mjs) and gh commands, with pinned models like sonnet, haiku, or opus and effort levels from medium to high.
What prerequisites do the exo skills require?▼
A git repository with a GitHub remote, the gh CLI authenticated (gh auth status is checked), and Node.js to run the bundled .mjs scripts. Session continuity relies on the using-exo skill loading naming and ordering rules at session start, after a clear, or after compaction.
How does exo handle unfinished work and repository-specific corrections?▼
The handoff skill saves an unfinished session's state so a fresh session after a clear can resume without loss. The memory skill retains user corrections about the repository, reports what exo remembers, and forgets lines on request, while savings tracks what the read guard kept out of context.