Test And Merge
Dev self-merge pipeline — merge main into branch, equiv tests, CI wait, self-merge via /dev-self-merge. Run by devs, not a separate tester agent.
Agent skills for software development, debugging, code review, and implementation. Browse reusable SKILL.md agent workflows and prompt templates in this category. Each card links to when-to-use notes and setup steps.
Dev self-merge pipeline — merge main into branch, equiv tests, CI wait, self-merge via /dev-self-merge. Run by devs, not a separate tester agent.
Test a tg-cli change against the real Telegram before it ships — pick the live scenarios the change touches, get the owner's yes, run them through bin/tg in Saved Messages, the test group and the test dialog only, snapshot and restore what they change, record the shape of every answer and never its content. Use when asked to test live, check a branch on the real account, run the live scenarios, «проверить вживую», «прогнать живые сценарии».
Run bounded, spec-first implementation cycles for Treeship's workflow and graph proof layer. Use when continuing workflow declarations, checkpoint composition, run binding, evidence-derived paths, bounded loops, or conformance verification.
Use when another agent hands Claude work — a handoff, an A2A task, an envelope, a message saying "take this from agent://…" — or when someone asks Claude to prove who it is. Refuses foreign work until the sender proves live key control, and records the verify in the handoff.
Use when the task in front of you is to move dependency pins — packages in either stack, Rust crates, tools, SDKs, GitHub Action references, or container images — or to find out which of them are behind and whether any changed licence.
Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.
Current weather and forecasts with webfetch, falling back to wttr.in curl for locations, rain, temperature, travel planning.
Guidance for explicitly requested new tests and for updating or reviewing existing tests in the Remotion repository.
xurl CLI for authenticated X posts, replies, reads/search, DMs, media upload, followers, auth status, or raw v2 API calls.
Plan and produce Chinese WeChat articles and spoken-video scripts, from news-topic selection and drafting to HTML formatting, publishing steps, and updates to the workflow’s content records.
Add and live-prove a model provider with non-interactive config one-liners, without exposing credentials.
Verify a named artifact, workflow, deployment, or script against observable evidence. Read-only by default, bounded to the requested target, and explicit when evidence is incomplete.
Use Better Stack API credentials bound in Nuphos to inspect or manage Uptime monitors and Telemetry resources.
Spend against a budget below the real window, and count the tool schemas you re-send on every step — they are the floor compaction cannot reach.
Bake, select, prove, and safely retire a Cloud Worker image with crabbox and config one-liners.
Use Cloudflare API credentials bound in Nuphos to inspect or change Cloudflare account resources.
Error-handling and resilience code review lens for catching silent failures, unbounded I/O, swallowed exceptions, and partial-failure pitfalls.
Configure and prove a chat channel with non-interactive one-liners; secrets only as SecretRefs.
Hill-climb the benchmarks in elizaOS/asi, a JAX continual-reinforcement-learning framework, with optional public payout registration. Use when an agent is asked to improve a measured result, port a method from the literature, fix something that corrupts measurement, decisively close a research direction with paired evidence, or register a public Solana payout address.
Crabbox and Blacksmith Testbox remote testing: isolation, cross-platform E2E, diagnostics, cleanup.
Create a new issue file from a test262 failure pattern. Includes smoke test, sample extraction, and proper frontmatter.
Measure whether a Topping Out rule actually binds, by tempting it at three prompt strictness levels in a throwaway working directory and grading what the session executed, deterministically. Use for /f-comply, "does this rule actually work", "test the NEVERs", "is RULES.md being followed", before trusting a new builder, or before promoting a rule to a hook. Measures existing rules: not for writing them (f-distill promotes a card into the constitution). Not a code gate.
Validate one plan against the codebase with COLD seats (the owner's standing choice, or the one-shot picker the first time and on request: 1-5 models from families different than the Player that authored the plan (codex/Astra, Claude, Grok CLIs at $0 subscription; OpenRouter doors from a curated roster), so the read is genuinely independent. Every seat runs at high reasoning, dispatched in parallel, and findings are merged across seats with seat attribution, never voted. Read-only but for two residues: the saved seat choice in the repo profile (the first pick, or a change) and the gate: plancheck line, written on request and always for a plan under or headed for the repo's ready-to-build plans folder. Use when the user runs /f-plancheck or /cv, wants a plan checked before building, or asks for cross-validation. This is the cold-seat protocol, not the validator itself, it dispatches plan-validator; reach for that directly only when cross-family independence does not matter. Also answers
Break a decision-complete plan (a /f-blueprint BLUEPRINT, or any plan too big for one job) into a PROJECT: one ticket file per job, a dependency graph, and the first project board (BOARD.json + a generated board page). Use when the user runs /f-schedule, says "break this plan into tickets", "decompose this into jobs", "turn this plan into a project", or when a plan has more than one PR's worth of work. Proposed-first: shows the cut and waits for the owner's go before writing. Not for writing the plan (f-blueprint), not for planning one ticket's job (the foreman's scoping + PLAN.md), not for board updates after that (f-board). Also answers to its old name, /f-decompose.