8
public repos
Agent skills · 278see the collection →
enforcing-context-budgets
Use when planned context, tool schemas, model output reserve, or retrieved artifacts may exceed tenant, route, model, or skill-specific token limits.
AI & Agents · casioreview20-glitch/forge-os
enforcing-review-separation
Use when artifact, code, finding, provider, or release approval requires independent review across principal, role, team, or trust-domain boundaries.
Code Review & Quality · casioreview20-glitch/forge-os
detecting-semantic-duplication
Use when several skills or providers may express the same trigger, mechanism, outputs, evaluation surface, or benchmark coverage under different names.
Testing & QA · casioreview20-glitch/forge-os
distilling-tool-output
Use when test, compiler, browser, database, or terminal output is too large for model context but must remain retrievable and evidentially complete.
Testing & QA · casioreview20-glitch/forge-os
binding-approvals-to-actions
Use when an irreversible, high-risk, scope-changing, or trust-changing operation requires a human decision that cannot be replayed for another payload or state.
Other · casioreview20-glitch/forge-os
calibrating-token-estimates
Use when estimated token counts differ systematically from provider-reported usage across models, content classes, languages, or tool schemas.
Content & Marketing · casioreview20-glitch/forge-os
certifying-release-evidence
Use when a source archive, package, container, skill registry, or product release needs reproducible proof bound to the exact bytes being distributed.
AI & Agents · casioreview20-glitch/forge-os
compiling-global-context
Use when an agent request must combine system policy, task state, skill sections, code, artifacts, memory, tools, and references under one hard model-context budget.
Workflow & Productivity · casioreview20-glitch/forge-os
composing-technique-workflows
Use when one outcome requires multiple specialized techniques with branch, join, parallel, retry, or rollback relationships rather than a single linear skill invocation.
AI & Agents · casioreview20-glitch/forge-os
consolidating-project-memory
Use when decisions, blockers, conventions, and stable summaries must survive across tasks without treating conversation transcript as the project source of truth.
AI & Agents · casioreview20-glitch/forge-os
decomposing-multi-domain-work
Use when one request spans independent product, engineering, design, security, data, legal, or operational subsystems that cannot be specified and verified as one bounded change.
Security · casioreview20-glitch/forge-os
expiring-stale-skills
Use when provider sources, tool APIs, model behavior, security assumptions, evaluators, or compatibility windows may have become outdated since the last successful validation.
Security · casioreview20-glitch/forge-os
Roles are derived from products, launches and repositories we track — not from self-description.