Hub Skill

1---
2name: agentclash-hub
3description: Use when starting any AgentClash eval, CLI, or challenge-pack task. Load this skill first for the full workflow map, skill dependency order, product UI links, hosted defaults, and pointers to every other AgentClash skill.
4metadata:
5  agentclash.role: hub
6  agentclash.version: "1"
7  agentclash.requires_cli: "false"
8---
9
10# AgentClash Hub
11
12## Purpose
13Give coding agents maximum context to run AgentClash evals through the CLI and guide humans to the right web UI pages — without reading the AgentClash source repository.
14
15## Use When
16- A user asks to evaluate agents, run evals, compare models, or use AgentClash for the first time.
17- You need to pick the right downstream skill before acting.
18- You need hosted defaults, UI links, or the end-to-end eval workflow in one place.
19
20## Do Not Use When
21- A narrower skill already matches (e.g. only CLI auth repair → `agentclash-cli-setup`).
22- The task is only to edit AgentClash product source code in the monorepo.
23
24## Environment
25Use production unless the user explicitly runs a local stack:
26
27```bash
28export AGENTCLASH_API_URL="https://api.agentclash.dev"
29agentclash auth login --device
30agentclash link
31agentclash quickstart
32```
33
34Install the CLI: `npm i -g agentclash` or see `/docs-md/getting-started/quickstart`.
35
36Portable bundle install (copy skills to agent host): https://github.com/agentclash/agent-skills
37
38## Procedure
391. Load this hub to pick the next skill.
402. Run `agentclash quickstart` if CLI readiness is unknown.
413. Follow dependency order for setup → pack → run → review → regression → CI.
424. Send the user to the matching UI page when they need a visual surface.
43
44## End-To-End Eval Workflow (CLI)
45
46```text
471. agentclash-cli-setup              → auth, workspace, doctor
482. agentclash-quickstart             → readiness checks + next command
493. agentclash-runtime-resources-setup → provider, model alias, runtime profile, secrets
504. agentclash-agent-build-author     → build spec + ready build version
515. agentclash-agent-deployment-setup → deployment ID for runs
526. challenge-pack skills             → plan, YAML, validate, publish pack
537. agentclash-eval-runner            → eval start / run create / follow / sessions / series
548. agentclash-scorecard-reader       → rankings, scorecards, replay, artifacts
559. agentclash-compare-and-triage     → baseline, compare latest/gate, replay triage
5610. agentclash-regression-flywheel   → promote failures, suite-only reruns
5711. agentclash-ci-release-gate       → CI manifest + gate (optional)
58```
59
60Optional branches (load when the workflow applies):
61
62```text
63• agentclash-multi-turn-operator     → human takeover during multi_turn runs
64• agentclash-dataset-workflows       → dataset eval, gate, traces, regression sync
65• agentclash-prompt-eval-playground  → prompt-eval YAML + playground experiments
66• agentclash-agent-harness-setup     → E2B coding-agent harness tasks and suites
67• agentclash-workspace-admin         → org/workspace CRUD and membership (teams)
68• agentclash-security-evaluation     → client-side security stress harnesses
69```
70
71Human-friendly shortcut after setup:
72
73```bash
74agentclash quickstart
75agentclash eval start --follow
76agentclash baseline set
77agentclash eval scorecard
78agentclash compare latest --gate
79agentclash replay triage
80```
81
82## Skill Dependency Order
83Read skills in this order when multiple apply:
84
851. `agentclash-hub` (this file)
862. `agentclash-skill-catalog` (when authoring or changing skills)
873. `agentclash-cli-setup`
884. `agentclash-quickstart`
895. `agentclash-runtime-resources-setup`
906. `agentclash-agent-build-author`
917. `agentclash-agent-deployment-setup`
928. `agentclash-challenge-pack-planner`
939. `agentclash-challenge-pack-yaml-author`
9410. `agentclash-challenge-pack-input-sets`
9511. `agentclash-challenge-pack-tools-sandbox`
9612. `agentclash-challenge-pack-artifacts`
9713. `agentclash-challenge-pack-scoring-validators`
9814. `agentclash-challenge-pack-llm-judges`
9915. `agentclash-challenge-pack-validation-publish`
10016. `agentclash-eval-runner`
10117. `agentclash-scorecard-reader`
10218. `agentclash-compare-and-triage`
10319. `agentclash-regression-flywheel`
10420. `agentclash-ci-release-gate`
10521. `agentclash-agent-harness-setup`
10622. `agentclash-multi-turn-operator`
10723. `agentclash-dataset-workflows`
10824. `agentclash-prompt-eval-playground`
10925. `agentclash-workspace-admin`
11026. `agentclash-security-evaluation`
111
112Each skill folder name matches its `name` in frontmatter. When a skill lists **Related Skills**, load those before mutating remote state.
113
114## All Skills In The Catalog
115
116| Skill folder | When to load |
117| --- | --- |
118| `agentclash-hub` | First — workflow map and UI links |
119| `agentclash-quickstart` | Readiness checks and suggested next command |
120| `agentclash-cli-setup` | Install, auth, workspace, doctor |
121| `agentclash-runtime-resources-setup` | Provider accounts, models, runtime profiles, secrets |
122| `agentclash-agent-build-author` | Agent build specs and build versions |
123| `agentclash-agent-deployment-setup` | Create/select deployments |
124| `agentclash-challenge-pack-planner` | Plan a pack before YAML |
125| `agentclash-challenge-pack-yaml-author` | Write pack YAML |
126| `agentclash-challenge-pack-input-sets` | Cases and input sets |
127| `agentclash-challenge-pack-tools-sandbox` | Tools and sandbox policy |
128| `agentclash-challenge-pack-artifacts` | Assets and artifact refs |
129| `agentclash-challenge-pack-scoring-validators` | Validators |
130| `agentclash-challenge-pack-llm-judges` | LLM judges |
131| `agentclash-challenge-pack-validation-publish` | Validate and publish |
132| `agentclash-eval-runner` | Start and follow evals, sessions, series |
133| `agentclash-scorecard-reader` | Interpret results |
134| `agentclash-compare-and-triage` | Baselines, compare, replay triage |
135| `agentclash-regression-flywheel` | Promote failures to regression suites |
136| `agentclash-ci-release-gate` | CI/CD gates |
137| `agentclash-agent-harness-setup` | E2B coding-agent harness tasks, suites, failure review |
138| `agentclash-multi-turn-operator` | Human takeover turns in multi_turn packs |
139| `agentclash-dataset-workflows` | Dataset eval, CI gate, traces, regression sync |
140| `agentclash-prompt-eval-playground` | Prompt eval YAML and playground experiments |
141| `agentclash-workspace-admin` | Org/workspace CRUD and membership administration |
142| `agentclash-security-evaluation` | Security pack stress-run and vault harnesses |
143
144Nested folders: `agent-build-skills/` and `challenge-pack-skills/` mirror the table rows above.
145
146## Product UI — Where To Send The User
147
148Base URL: **https://agentclash.dev**
149
150| User goal | UI path |
151| --- | --- |
152| Sign in / account | https://agentclash.dev |
153| Docs home | https://agentclash.dev/docs |
154| Quickstart | https://agentclash.dev/docs/getting-started/quickstart |
155| First eval walkthrough | https://agentclash.dev/docs/getting-started/first-eval |
156| Agent skills (web catalog) | https://agentclash.dev/docs/agent-skills |
157| CLI reference | https://agentclash.dev/docs/reference/cli |
158| Challenge packs guide | https://agentclash.dev/docs/guides/write-a-challenge-pack |
159| Multi-turn packs | https://agentclash.dev/docs/challenge-packs/multi-turn |
160| Interpret results | https://agentclash.dev/docs/guides/interpret-results |
161| CI/CD gates | https://agentclash.dev/docs/guides/ci-cd-agent-gates |
162| Workspace runs (after login) | App dashboard → Runs list |
163| Live run events | Run detail page while status is running |
164| Scorecards & comparisons | Run detail → scorecard / ranking views after completion |
165
166When you create a run via CLI, tell the user:
167
168```text
169Open https://agentclash.dev and navigate to your workspace runs, or search for run ID <RUN_ID> after signing in.
170```
171
172## AgentClash Concepts (30-Second Model)
173
174- **Challenge pack** — versioned eval workload (cases, scoring, tools policy).
175- **Input set** — which cases run in a given eval.
176- **Agent build / deployment** — the agent under test (model + runtime + tools).
177- **Run** — one execution of pack × input set × deployments.
178- **Eval session** — repeated runs (`eval start --repetitions N` or `run series create`).
179- **Scorecard** — structured results, comparisons, release gate input.
180- **Baseline bookmark** — workspace default run/agent for `compare latest`.
181- **Regression suite** — promoted failures for suite-only reruns.
182
183## Expected Output
184After loading this skill you can name the next skill, 1–3 CLI commands, and the UI page the human should open.
185
186## Failure Modes
187- Skipping `agentclash-cli-setup` when auth or workspace is unset → commands fail with workspace errors.
188- Running evals before pack publish → no runnable pack version.
189- Using localhost API URL by mistake → empty workspace or auth failures against the wrong backend.
190
191## Safety Notes
192- Confirm before production-scale evals, publishes, or CI runs that spend budget.
193- Never paste tokens, secrets, or customer data into chat.
194- Prefer `agentclash doctor` and read-only list commands before writes.
195
196## Report Back Format
197```text
198Hub loaded: yes
199Next skill: <skill-folder-name>
200CLI status: <auth/workspace/doctor summary>
201UI for user: <https://agentclash.dev/...>
202Next commands: <1-3 commands>
203```
204
205## Related Skills
206Load all skills listed in **Skill Dependency Order** as needed; start with `agentclash-cli-setup` if CLI is not configured.
207
208## Related Docs
209- `/docs-md/agent-skills`
210- `/docs-md/agent-skills/agentclash-hub`
211- `/docs-md/guides/use-with-ai-tools`
212- `/docs-md/getting-started/quickstart`
213- `/docs-md/getting-started/first-eval`