Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/Pierry/harness-kitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/pierry/harness-kit/firebase-publish)<a href="https://agentmods.dev/commands/pierry/harness-kit/firebase-publish"><img src="https://agentmods.dev/badge/commands/pierry/harness-kit/firebase-publish/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/pierry/harness-kit/firebase-publish"><img src="https://agentmods.dev/badge/commands/pierry/harness-kit/firebase-publish.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00020 | $0.00971 |
| Opus 5.5 | $0.00008 | $0.00388 |
| Sonnet 5.5 | $0.00004 | $0.00194 |
| Haiku 4.5 | $0.00002 | $0.00097 |
Grade A, and why
firebase-publish scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
GOTCHA: when `public` is the repo root, `**/.*` does NOT exclude a dotdir's non-dot children, so `.claude/**` would leak PRDs/PRPs/internals publicly. Always ignore `.claude/**`, `.git/**`, and `*.md` explicitly. After d How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Publish static site to Firebase Hosting. Create project if missing, else reuse. Deploy.
Print header card before deploy, footer card after live URL returns. Format: .claude/scripts/stage-card.md.
Hard safety rules
- NEVER delete. No
projects:delete, nohosting:sites:delete, nohosting:channel:delete, no resource removal. If wrong project made, tell user to delete in console. Do not attempt. - Create AT MOST ONE project per run. After
projects:create, re-checkfirebase projects:list. Confirm exists. Do NOT loop create on ambiguous output. CLI printsYour Firebase project is ready!on success, notSuccess. Match loose, verify by list, stop after one. - Outward-facing publish. Live site = public. Confirm PROJECT_ID with user before create unless user passed it as arg.
Inputs
- PROJECT_ID. user arg, else derive from feature_id/repo slug, propose, confirm. Globally unique, lowercase, 6-30 chars.
- SITE_ID. default = PROJECT_ID.
Preconditions
firebase --versionexists. Missing → tell usernpm i -g firebase-tools. Stop.firebase login:list. Not logged in → tell user run! firebase login. Stop. Never write creds.
Steps
- Resolve PROJECT_ID (above).
firebase projects:list. PROJECT_ID present → reuse, skip create. Absent →firebase projects:create <PROJECT_ID> --display-name "<name>". Then re-runprojects:list, confirm present. Not present → stop, surface error, ask user.- Pin default: write
.firebaserc{"projects":{"default":"<PROJECT_ID>"}}if absent or different. - Ensure
firebase.jsonhosting block exists. Missing → write minimal:
Confirm{ "hosting": { "public": ".", "ignore": ["firebase.json", ".firebaserc", ".firebase/**", ".git/**", "**/.git/**", ".gitignore", ".claude/**", "**/.claude/**", "**/node_modules/**", "*.md"], "cleanUrls": true } }publicpoints at the dir holdingindex.html. GOTCHA: whenpublicis the repo root,**/.*does NOT exclude a dotdir's non-dot children, so.claude/**would leak PRDs/PRPs/internals publicly. Always ignore.claude/**,.git/**, and*.mdexplicitly. After deploy, verify the file count is small andcurl -o /dev/null -w '%{http_code}' <url>/.claude/...returns 404. - Stage first:
firebase hosting:channel:deploy preview --project <PROJECT_ID> --expires 7d. Capture preview URL. Show user. - Promote:
firebase deploy --only hosting --project <PROJECT_ID>. - Live URLs:
https://<SITE_ID>.web.appandhttps://<SITE_ID>.firebaseapp.com. - Verify no leak: deploy log file count should match the site (small).
curl -s -o /dev/null -w '%{http_code}' https://<SITE_ID>.web.app/.claude/...must return 404, home must return 200. Same check on the preview channel URL. Leak found → fixignore, redeploy live AND preview.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 67 lines · 20 tokens per session scan A 3197f7f6cf44
firebase-publish is a command published in the GitHub repository Pierry/harness-kit (4 stars, last pushed 13d ago), licensed MIT. It adds 20 tokens to every session and 971 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
p7-deploy
Executes the deployment to the target environment and then verifies with smoke tests that the system is running correctly. Only when the smoke tests pass is the deployment considered successful.
p7-monitoring
Sets up production monitoring completely, configures alerting rules, and tests the rollback plan. The goal is a system that reports problems early and where a rollback works smoothly in an emergency.
p7-prepare
Prepares the deployment: checks all environments, prepares database migrations, verifies configurations, and ensures nothing has been overlooked. Only after this checklist is complete will /p7-deploy be executed.
p3-infra-hosting
Plans hosting platform, deployment model, environments and CDN strategy.
p6-audit-config
Checks security headers, CORS, DB access, secrets management, and server configuration.
setup-cdn-optimization
Command "setup-cdn-optimization" from qdhenry/Claude-Command-Suite, covering setup cdn optimization and instructions.