Pubx-ai

@pubx-ai source repo

15 skills
0 followers
14 hours ago last updated

▌
Raise Pr · pubx-ai
Push the current feature branch and raise a GitHub pull request with a conventional-commit style title and a structured body, after verifying the branch actually has committed work to raise. Use this skill whenever the user asks to "raise a PR", "open a pull request", "create a PR", "push and PR", "get this reviewed", or wants the current branch turned into a pull request. This is the last step of the delivery flow — after create-branch and one or more create-commit commits — and hands off to address-pr-comments once reviews arrive.

0
installs
▌
Run Tests · pubx-ai bundle
Check whether the current project has a test suite, work out the correct way to run it, run it, and report results faithfully. Use this skill whenever the user asks to "run the tests", "do the tests pass", "run the test suite", "check the tests", "test this", "run pytest / vitest / bun test / playwright", or when a change needs test verification before commit or PR. Detects the runner from the project's own instructions (AGENTS.md, CLAUDE.md, .cursor/rules), Makefile/package scripts, and lockfiles rather than guessing from memory. Also handles projects that intentionally have no tests. DO NOT USE for writing new tests, designing test strategy, or reviewing coverage (use local-test-review for those), CI pipeline configuration, or debugging test logic (use systematic-debugging for that).

0
installs
▌
Create Plan · pubx-ai bundle
Turn a completed requirements document into a detailed, sequential implementation plan that an AI coding agent can execute autonomously. Use this skill whenever the user asks to "create an implementation plan", "plan this feature", "break this down into tasks", "generate a task breakdown", "plan the implementation of…", or points at a requirements doc (e.g. under docs/requirements/) and asks what to build next. This is the natural next step after the gather-requirements skill — reach for it as soon as requirements are signed off. Also use it when handed a spec or detailed ticket where the ask is to sequence the work into executable tasks rather than to start coding immediately.

0
installs
▌
Adcp Review · pubx-ai bundle
Review code changes for conformance with the AdCP (Ad Context Protocol, agenticadvertising.org) technical spec. Use whenever reviewing a PR, diff, commit, or code change in a repository that implements AdCP — seller agents, buyer agents, orchestrators, or SDK integrations — even if the request is just "review this PR" or "check my changes" without mentioning AdCP explicitly. Covers task request/response wire shapes, async task lifecycle and status handling, idempotency, error envelopes, authentication/signing, webhook security, and media-buy semantics. Also use when asked whether a change "breaks the protocol", "is spec-compliant", or "will interop with other AdCP agents". DO NOT USE for reviewing proposals to change the protocol itself — PRs or branches in the adcontextprotocol/adcp spec repository — where the live docs are the baseline being changed, not the contract; use adcp-proposal-review instead.

0
installs
▌
Execute Plan · pubx-ai
Execute a written implementation plan task by task, with a critical review of the plan before starting and a checkpoint after every task. Use this skill whenever the user asks to "execute the plan", "implement the plan", "run the plan", "work through the plan", "start on task 1", "continue the plan", or points at a plan directory or file (e.g. under docs/plans/) and asks to build it. This is the natural next step after the create-plan skill — reach for it as soon as a plan exists and the ask is to start coding. If a skill named "executing-plans" is available in the session, this skill delegates the loop to it rather than duplicating it, while keeping its own review-gate and checkpoint invariants.

0
installs
▌
Local Review · pubx-ai
Act as a senior engineer and review all uncommitted changes in the local git branch — both staged and unstaged, including new untracked files — before they get committed. Use this skill whenever the user wants pre-commit feedback on work in progress — phrases like "review my changes", "review my diff", "code review before I commit", "look over what I changed", "is this ready to commit?", "check my local changes", or asking for feedback on the correctness, security, performance, tests, or style of uncommitted work. Pairs with the create-commit skill: review first, commit after. Produces feedback anchored to file paths and line numbers plus a 0–10 score and an APPROVE / REQUEST_CHANGES style recommendation.

0
installs
▌
Create Branch · pubx-ai
Create a correctly named git feature branch off an up-to-date base, carrying the JIRA ticket ID when one is in context. Use this skill whenever the user wants to "create a branch", "make a new branch", "branch off main", "start work on PUB-1234", "checkout a feature branch", or when work is about to begin (or a commit is about to land) while still on main/master. This is the first step of the delivery flow — branch, then commit with create-commit, then raise the PR with raise-pr.

0
installs
▌
Create Commit · pubx-ai
Compose and create a git commit whose message follows the Angular Conventional Commits standard, so semantic-release can version and changelog it automatically. Use this skill whenever the user wants to commit staged changes, "make a commit", "commit this", "write a commit message", "commit with conventional/angular format", or asks what type/scope a change should be. Also use it when a ticket ID (e.g. PUB-4204) should be pulled from the branch name into the commit footer. The skill reviews the staged diff first and only commits after the change has been reviewed.

0
installs
▌
Adcp Scorecard · pubx-ai bundle
Evaluate whether AdCP (Ad Context Protocol, agenticadvertising.org) — or a concrete AdCP implementation such as a seller agent, buyer agent, or orchestrator — is safe, interoperable, operationally viable, and ready for pilot or production use. Combines a weighted decision scorecard, hard safety gates that override any numerical score, and an observable pilot test suite run against a sandboxed or live agent. Use when asked to "score AdCP", "evaluate AdCP", "run the AdCP scorecard", "is AdCP (or our seller agent) production-ready", "pilot-test the AdCP agent", "assess AdCP readiness", or to compare AdCP against another agent protocol with a weighted rubric. DO NOT USE for reviewing code changes for AdCP spec conformance (use adcp-review) or for routine operations against a seller agent — discovery, media buys, updates, deliveries (use adcp-seller-agent).

0
installs
▌
Local Pr Review · pubx-ai
Act as a senior engineer and review the committed changes on the current git branch against the base branch (develop or main) — a local, pre-push PR review. Use this skill whenever the user wants feedback on a branch as a whole before opening or merging a pull request — phrases like "review my branch", "PR review before I push", "review my changes against develop/main", "is this branch ready to merge?", "review the whole feature branch", or asking about the correctness, security, performance, tests, or style of the commits that diverge from the base branch. Unlike local-review (which looks at uncommitted work), this reviews the branch's committed diff. Produces feedback anchored to file paths and line numbers plus a 0–10 score and an APPROVE / REQUEST_CHANGES style recommendation.

0
installs
▌
Adtech Expert · pubx-ai bundle
Answer AdTech questions as a practitioner with Ad Ops and Engineering expertise: programmatic advertising, header bidding and the Prebid ecosystem, Google Ad Manager and other ad servers, OpenRTB and IAB specs, agentic advertising and AdCP concepts, audience and data ingestion pipelines, cleanrooms, and adtech API integrations. Use whenever the user asks how something works, why something behaves oddly, or how to build/integrate/debug anything in this space — "why is my bid not winning", "how do price granularity buckets work", "design an audience ingestion pipeline", "GAM line item priority vs header bidding", "compare cleanroom approaches" — even when they don't say "adtech". Conceptual AdCP and agentic advertising questions are in scope; hands-on protocol operations (driving live agents, conformance review, readiness evaluation, spec proposals) are not — this skill answers and designs, it does not operate.

0
installs
▌
Local Bot Review · pubx-ai
Run a CodeRabbit CLI review over local changes — a review bot's independent second opinion before anything is pushed. Use this skill whenever the user asks to "run a bot review", "run coderabbit locally", "coderabbit review my changes", "get a machine review", "run the review bot on this diff", or wants bot findings on uncommitted work, a branch, or a commit range without opening a PR. Verifies its own prerequisites (CodeRabbit CLI installed, authenticated, version ≥ 0.4.0) and delegates the review mechanics to the CodeRabbit "code-review" skill when it is installed, recommending its installer when it isn't. DO NOT USE for triaging comments on an open PR (use address-pr-comments) or for agent-reasoned review (local-review for uncommitted work, local-pr-review for a branch) — this skill is the machine's eyes, not the agent's.

0
installs
▌
Local Test Review · pubx-ai
Review a project's testing state against a proper test strategy: inventory what tests exist, baseline what passes, identify the gaps that matter, and produce a prioritized test plan — running the suite via the run-tests skill when one exists. Use this skill whenever the user asks "how should we test this", "review our test coverage", "what tests do we need", "is our testing adequate", "test strategy for X", "assess the tests", or wants a test plan for a feature or repo. If a skill named "testing-strategy" is available in the session, this skill delegates the strategy design to it. DO NOT USE when the user just wants the tests executed — that's run-tests — or wants a specific test written (that's implementation work).

0
installs
▌
Address Pr Comments · pubx-ai bundle
Load every comment on a pull request, understand and triage each one against the current code, then resolve them — fixing what's valid, pushing back with technical reasoning on what isn't. Use this skill whenever the user asks to "address the PR comments", "review the PR comments", "respond to review feedback", "handle the CodeRabbit / Bugbot / reviewer comments", "triage the review comments", "fix the PR feedback", or "resolve the comments" on a pull request. Works for human and bot reviewers alike. If a skill named "receiving-code-review" is available in the session, this skill applies its verification discipline rather than duplicating it: verify before implementing, no performative agreement, technical correctness over social comfort.

0
installs
▌
Gather Requirements · pubx-ai bundle
Turn a raw, underspecified feature request into a structured, implementation-ready requirements document before any planning or coding begins. Use this skill whenever the user describes something they want built but the details are thin — phrases like "I want to add X", "we need a feature that…", "gather requirements", "flesh out this ticket", "turn this into a spec", "write the requirements for…", or "what do you need to know before building this?" — and also when handed a rough Jira ticket, user story, or one-liner that lacks technical context, acceptance criteria, or scope. Also use it when the user wants an interview-style interrogation of a feature idea — "grill me about this feature", "stress-test these requirements", "interview me about what I want". It reads the project's own conventions, asks targeted clarifying questions one decision at a time with a recommended answer, and writes a complete requirements doc to a file so an implementation planner has everything it needs. Prefer this over jumping str

0
installs

15 published skills