Profile My skills Saved Collections Edit profile Submit a skill

← all publishers

agentic-tend

@agentic-tend source repo

11 published skills

  1. ▌
    Obsidian CLI · agentic-tend bundle
    Use the Obsidian CLI when a task depends on Obsidian's index or running application state, such as resolved links or properties, Bases, or link-preserving moves. Do not use it for ordinary filesystem inspection, Markdown or YAML editing, Git work, or plugin and theme development that does not need Obsidian runtime semantics.
    0
    installs
  2. ▌
    Referee Response · agentic-tend bundle
    Draft or revise a response to an exact scientific referee comment using the current manuscript and already established evidence, and map promised manuscript changes. Use when response-to-reviewers prose is the direct deliverable. It consumes rather than decides contested scientific validity; do not use for ordinary manuscript editing or general journal correspondence.
    0
    installs
  3. ▌
    Julia Development · agentic-tend bundle
    Own Julia-specific decisions about semantics, dispatch, API design, and host conventions for code, comments, docstrings, and Documenter.jl using the target project's contracts and version-compatible official Julia documentation. Use for any Julia implementation, refactor, review, diagnosis, or Documenter.jl realization, and for performance work when requested or when evidence identifies a hot path, allocation, type-instability, benchmark regression, threading issue, or new performance promise. Do not own language-independent software topology or prose organization.
    0
    installs
  4. ▌
    Markdown Authoring · agentic-tend bundle
    Own decisions about Markdown source syntax and generic or GitHub renderer realization when work changes Markdown files, links, headings, tables, footnotes, disclosures, diagrams, mathematics, source formatting, or renderer compatibility. Compose with $structure-documentation for reading-path and prose decisions and with a host capability for renderer extensions such as Documenter.jl. Do not decide prose meaning or operate application runtime state.
    0
    installs
  5. ▌
    Clarifying Contracts · agentic-tend bundle
    Resolve user-authoritative semantic ambiguity after inspection identifies an unresolved choice that could change purpose, direct object, public behavior, constraints, invariants, scope, decision authority, or acceptance evidence, or when the user asks to clarify or stress-test such a contract. Do not use for discoverable facts or agent-owned implementation choices that preserve the contract.
    0
    installs
  6. ▌
    Multi Agent Evidence · agentic-tend bundle
    Use multiple agent contexts when independent evidence, held-out checks, or posterior falsification can reduce correlated failure or human review load. Expose a plan before delegation and a provenance-aware delivery. Do not use for agent count, importance, task size, latency-only parallelism, repeated opinions, stylistic review, or majority voting.
    0
    installs
  7. ▌
    Software Engineering · agentic-tend bundle
    Own software-engineering decisions about behavior, topology, implementation, testing, feedback, refactoring, performance, review, delivery, task-local structured IR, and whether embedded prose must preserve non-inferable software meaning. Use for software work and when pseudocode is requested for software logic, data, state, concurrency, lifecycle, or recovery. Do not use for prose-only expression or host-language attachment that makes no software decision.
    0
    installs
  8. ▌
    Scientific Claim Audit · agentic-tend bundle
    Evaluate whether a specific scientific claim or scientific-validity objection is supported by controlling derivation, data, code, figure, manuscript context, or cited-source evidence, and return its support status, scope, and unresolved gaps. Use for claim-validity review, citation-support checks, and claim-level inconsistencies across scientific artifacts. Do not use merely because content is scientific, or for explanation, copyediting, style or notation review, literature discovery, response drafting, or faithful format transfer.
    0
    installs
  9. ▌
    Structure Documentation · agentic-tend bundle
    Organize and express durable, intentionally authored natural language independently of artifact syntax. Use when persistent prose is the direct task object or another capability has established prose to write, including documentation and natural-language spans inside code or interfaces. Do not decide whether software prose is needed, what semantic contract it owns, or which artifact or host-language syntax is valid; do not apply directly to formal code or data, generated content, transient chat, or verbatim quotations.
    0
    installs
  10. ▌
    Bootstrap Project Context · agentic-tend bundle
    Derive, audit, or migrate persistent repository agent context and its decision owners from durable local evidence, then apply the minimum sufficient result, including a valid zero-file result. Use for repository agent rules, decisions, local skills, mechanical enforcement, generators, or retirement of generic agent scaffolding. Do not use for ordinary implementation or transient planning.
    0
    installs
  11. ▌
    Scientific Literature Evidence · agentic-tend bundle
    Build and evaluate a research-literature corpus for a scientific landscape, attribution or priority question, version comparison, source conflict, or bounded absence search. Return source roles, claim-level evidence, version state, and search limits. Do not use for generic web search, bibliography formatting, or checking one already-supplied citation.
    0
    installs