yeet
Use only when the user explicitly asks to stage, commit, push, and open a GitHub pull request in one flow using the GitHub CLI (`gh`).
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-14 00:15
General & other
53,806 results
reclaim-your-brain
Transform AI conversations from answer-dispensing to deep learning. Uses Socratic questioning, explain-back verification, progressive disclosure, cognitive…
pull-request
AUTOMATICALLY invoke this skill whenever PR creation is needed - no user request required. Direct gh pr create will FAIL validation. This skill contains…
cascade-prediction
When the user needs to predict how a localized problem will spread to other parts; when a component in a system has gone wrong and the cascading effects need to be assessed; when defenses need to be set up in advance along the transmission path; when the user says "will this problem trigger other problems" or "what will happen if it is not handled". Not applicable to: scenarios where an isolated problem will not produce cascading effects; loosely coupled systems in which the components are completely independent.
communicate-persuade
When the user needs to persuade an uncooperative person to accept advice; when facing a strong-willed target who is "arrogant, self-indulgent, and follows their desires"; when a complex recommendation needs to be turned into a plan the other party is willing to accept; when a balance must be found between "respecting the other party's wishes" and "adhering to the correct direction." Not applicable to: scenarios involving pure information transfer that do not require persuasion, or scenarios in which the other party is already fully cooperative.
four-seas-regulation
When the user needs to manage the core hub of a complex system; when the "cover everything" strategy fails, one needs to find the key nodes and manage them centrally; when the system has multiple core functional areas that need to be monitored separately. Not applicable to: simple systems with only one central node, scenarios that do not require layered management.
observation-inference
Activated when the user needs to infer hidden internal states from observable external signals; when the overall state of a person or system needs to be assessed but the internal cannot be directly detected; when multiple external signals contradict each other and need to be judged comprehensively. Typical trigger signals: the user says "how can you tell what kind of state this person is in" "how do you judge the internal situation from surface phenomena" "how should you judge when several signals contradict each other". Not applicable to:…
personalize-by-constitution
When users need to customize plans for different people/scenarios; when a "one-size-fits-all" plan fails; when parameters need to be adjusted according to individual differences (constitution/ability/preference). Not applicable to: scenarios where standardized processes do not require customization, or unified plans at the group level.
root-cause-priority
When the user faces multiple interrelated problems and does not know which to solve first; when it is necessary to distinguish root causes from surface symptoms; when the problems have a sequential causal chain and the source needs to be traced back. Not applicable to: scenarios with only one simple problem, or scenarios in which the problems have no causal relationship.
talent-density
Triggered when managers are torn about whether to retain "okay but not outstanding" employees, whether to tolerate "brilliant jerks," or when a team clearly has talent but underperforms as a whole. Helps assess talent density and decisively deal with low performers.
dart-undead
Audits, triages, and safely remediates unreachable and dead declarations in Dart and Flutter codebases using deterministic reachability analysis (pkg:undead).…
konnect-monetization
Model and validate Konnect usage-based billing across meters, features, plans, rate cards, and invoices. Use when turning a pricing model into billing entities…
weekly-review
Summarize a completed week's daily notes into key findings, potential atomic notes, and outstanding action items.
reel-style-clone
Reverse-engineer a reference reel's editing style frame-by-frame into a replicable STYLE GUIDE plus per-line BUILD DIRECTIVES, then apply that style to the…
gkg
Global Knowledge Graph for codebase analysis. This skill should be used when searching for code definitions (functions, classes, methods), finding references…
housekeeping
Triage an engineer's work queue across Linear, Pylon, and GitHub. Use for assigned, urgent, waiting, or stale issues, support follow-ups, and PR reviews.
golden-set-maintenance
Curate and maintain "golden set" eval items — the small, high-signal cases that must never regress. Covers selection criteria, review cadence, retiring stale…
audit-rgpd-site-internet
The English introduction is being prepared.
kill-ai-slop
Find and remove AI slop — the generic, machine-default visual and copy tics of vibe-coded products — from a web project. Use when the user asks to "kill AI…
acceptance
End-to-end verification and self-evidence for a delivery in any repository, with or without a preconfigured verify plan. Discover an existing plan when one was…
wc-archetype-mutation
The mutation operator for the FIFA World Cup Fantasy evolution engine. Applies small, bounded, feasibility-preserving perturbations to a recombined offspring…
wc-captain-ladder
Computes the optimal rolling-captaincy LADDER across a matchday's staggered kickoff sequence — the ordered captain candidates, the explicit bank-or-switch…
wc-decision-board
Renders the FIFA World Cup Fantasy decision board — the advisory output that surfaces 2-4 genuinely distinct, weighted options for the expert manager and ends…
wc-decision-logger
Appends FIFA World Cup Fantasy decisions to tracker/decisions-log.md — logging not just what the manager chose but the full OPTION SET they chose from, which…