mark-torres10
Community@mark-torres10
A personal registry of 30 agent skills for software engineering teams covering PR authoring, code review, planning, CI repair, and prose editing.
Agent Skills by mark-torres10
Showing 30 vetted skills indexed across 1 GitHub repositories.
mark-torres10write-pr-description
Writes type-specific PR descriptions for experiments, features, bug fixes, and default changes.
mark-torres10write-changelog
Writes terse, outcome-focused CHANGELOG entries for merged pull requests.
mark-torres10implement-epic
Implements a GitHub epic as stacked PRs, one PR per child issue via gh stack.
mark-torres10comprehensive-code-review
Conducts structured code reviews using a collaborative checklist covering per-file analysis and risks.
mark-torres10explain-as-python
Explains TypeScript and JavaScript code to Python developers through concept teaching and Python translation.
mark-torres10implement-plan-and-open-pr
Implements an existing plan end-to-end and opens a verified pull request.
mark-torres10plain-writing
Rewrites prose in a plain style with simple words, complete sentences, and no jargon.
mark-torres10review-persona
Reviews current work using a selected expert persona from an agents/personas directory.
mark-torres10implement-from-spec
Implements scoped code slices from approved design docs using phased test-first workflows.
mark-torres10interactive-implementation
Executes implementation plans step-by-step with docstring, simplicity review, and approval gates.
mark-torres10handoff
Generates a structured end-of-session handoff.md summarizing completed work, remaining to-dos, and PR status.
mark-torres10create-advisory-brief
Generates a self-contained markdown advisory brief for external AI evaluation of architecture and vendor decisions.
mark-torres10create-implementation-plan
Generates structured implementation plans with executive summaries and per-step detail files.
mark-torres10delegate-work-to-subagents
Orchestrates multi-agent workflows by delegating research, implementation, QA, and verification to subagents.
mark-torres10probe-understanding-professor
Quizzes users with adaptive professor-style questions and writes scored markdown understanding reports.
mark-torres10suggest-rules-additions
Analyzes conversation history to suggest additions to a docs/RULES.md preferences file.
mark-torres10probe-understanding-manager
Probes user understanding of engineering designs through manager-style questioning and scored markdown reports.
mark-torres10review-for-simplicity
Reviews plans, proposals, and code changes for unnecessary complexity and premature abstraction.
mark-torres10latex-exact-transcriber
Converts text or image content into LaTeX output using only dollar-sign delimiters.
mark-torres10fix-ci
Diagnoses failing GitHub PR checks, reproduces failures locally, fixes code, and pushes verified commits.
mark-torres10setup-new-repo
Bootstraps greenfield repositories with uv, pre-commit hooks, CI workflows, and GitHub remotes.
mark-torres10write-docstring
Writes numpy-style Python docstrings for functions, classes, and modules using routed reference guides.
mark-torres10create-epic
Creates a GitHub epic with parent and child sub-issues from an implementation plan.
mark-torres10review-security
Applies Semgrep code-security and OpenAI security-best-practices skills to review code for vulnerabilities.
Frequently Asked Questions About mark-torres10
FAQPage SchemaWhat tasks can I accomplish with mark-torres10's skills?▼
You can write type-specific PR descriptions and changelogs, create GitHub epics with child issues, implement plans end-to-end with stacked PRs, run comprehensive and persona-based code reviews, fix failing CI, write numpy-style docstrings, refactor services, and clean up prose with plain-writing and anti-slop editing.
Who are these skills designed for?▼
Software engineers and engineering managers working in GitHub-centric repositories. Specific personas include Python-fluent developers learning TypeScript, tech leads pressure-testing designs with the manager or professor probes, and writers or DevRel staff tightening technical prose and launch copy.
How do the execution skills run in practice?▼
Execution skills like implement-epic, implement-plan-and-open-pr, and fix-ci are slash-invoked or explicitly requested. They orchestrate steps such as spawning per-issue orchestrators, running write-docstring and review-for-simplicity gates between steps, committing, pushing, and returning the resulting PR URL.
Are these skills open source and what do they cost?▼
The anti-slop-writing and humanizer skills carry explicit MIT licenses and credit their upstream sources; grill-me is sourced from a public repository. Other skills list no license or fee. All are free to inspect in the registry, with no stated usage cost.
What prerequisites or dependencies do these skills require?▼
Skills assume an agent-skills-compatible client such as Codex, OpenCode, Gemini CLI, or Claude Code. GitHub skills require gh CLI access for stacked PRs and epics, review-security invokes Semgrep, and several skills depend on project files like docs/RULES.md, agents/personas, and types/ templates.