Observal is a self-hosted registry and control plane for an organization's internal AI components, including skills, agents, and MCP servers. It is for teams that need to discover, share, manage, and understand usage of coding-agent extensions across their organization. The catalogue entries are Observal's own skills and instructions for operating the registry and its insight engine.
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.
npx skills add Observal/Observal --skill observal-registrygit clone --depth 1 https://github.com/Observal/ObservalWrote 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/skills/observal/observal/observal-registry)<a href="https://agentmods.dev/skills/observal/observal/observal-registry"><img src="https://agentmods.dev/badge/skills/observal/observal/observal-registry/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/skills/observal/observal/observal-registry"><img src="https://agentmods.dev/badge/skills/observal/observal/observal-registry.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00000 | $0.00813 |
| Opus 5.5 | $0.00000 | $0.00325 |
| Sonnet 5.5 | $0.00000 | $0.00163 |
| Haiku 4.5 | $0.00000 | $0.00081 |
Grade A, and why
observal-registry scanned grade A with 0 findings 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 14d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Managing Registry Components
Execution contract
- Execute commands with a 60 second timeout.
- Use machine output by default: add
--output jsonwhenever supported. Parse list results fromitemsand pagination fields. - Run the leaf command's
--helpwhen any path, flag, enum, or payload shape is uncertain. - Supply all required inputs and confirmation flags. Do not leave an agent waiting at a prompt.
- Reuse returned UUIDs and
qualified_namevalues. Never automate with row numbers or ambiguous bare names. - Verify installs, submissions, edits, versions, ownership changes, and lifecycle transitions.
- Submit or modify only components the user owns or is authorized to manage.
- Never expose environment values, headers, tokens, private source data, or submitted secret fields.
- Mutations are sent once. After an uncertain transport failure, read component state before retrying.
- Authentication is optional for approved public content when the server setting
deployment.public_registry_enabledis enabled; it is disabled by default on self-hosted deployments. Public list, show, install, and prompt render commands usehttps://public.observal.ioby default. Authenticate before submitting, editing, reviewing, rating, or accessing private team content.
Choose the workflow
| User intent | Read |
|---|---|
| Find, inspect, recommend, or install components | Discovery and installation |
| Submit one component or a mixed bulk file, or register a remote A2A agent | Component submission |
| Fork, edit, version, archive, restore, transfer, or manage co-authors | Registry lifecycle |
Read only the selected reference, and read it completely before executing.
Registry rules
- Search with the user's natural-language terms, then narrow by type, namespace, team, harness, or category only when useful.
- Open-ended requests such as "what am I missing?" use personalized recommendations before keyword search.
personalized: falsemeans popularity fallback, not a personal recommendation.- Team members can see authorized private teamspace items. Use
TEAM_HANDLE/ITEM_SLUGfor direct references. - Draft, pending, rejected, and approved items have different edit behavior. Read status before mutating.
- Forks copy only currently approved sources and approved releases into independent drafts. Do not disclose source identity if the API reports
forked_from.available: false. - A successful submit can still be pending review. Report the returned status instead of saying it is published.
- Bulk files are structurally validated before mutation. Inspect every per-entry result and verify uncertain retries by canonical identity.
- Prefer an existing installed dependency or native CLI path. Do not invent wrappers or telemetry variables.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 14d ago Changed · +1 lines e8a9ff52fccf
- 27d ago Changed · +2 lines · -76 tokens per session 00536d64b93c
- 1mo ago Changed 7d7f5809b3f6
- 1mo ago Changed · +1 lines 1151ed7ab22f
- 1mo ago First seen · 49 lines · 76 tokens per session scan A 92347b015074
observal-registry is a skill published in the GitHub repository Observal/Observal (4,533 stars, last pushed yesterday), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 813 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
commonly
You are a member of a Commonly workspace — a shared space where humans and AI agents from any origin collaborate in pods (chat rooms with memory). Use this whenever you are connected to Commonly via the commonly MCP tools: to read what's happening, post, remember things across sessions, react, DM other agents, and…
pod-chat-tone
In-product voice rewrite for agents posting to Commonly pods — kills the report-in-a-chatbox shape (bold-lead sentences, ✅/❌ ladders, em-dash density, mirror clauses, self-narration of one's own reasoning) that makes agent rooms unreadable to the humans they are for. Mandatory pre-post pass on every pod message.…
pandic-office
Convert Markdown to PDF (or DOCX/EPUB/HTML) using the pandoc CLI. Use when asked to produce a PDF report, brief, summary, or any document where the input is Markdown and the output should be a polished, paginated file.
Manipulate PDF files — extract text, count pages, render thumbnails, merge or split documents. Use for PDF-specific operations that don't fit markdown-converter (general read) or pandic-office (write from markdown).
tmux
Manage long-running shell sessions with tmux — start a detached session, run a long task, reattach later, capture output. Use when a task takes longer than a single tool call (build, test, log tail).
commonly-design
Use this skill for ANY visual / UI / design work on Commonly — picking colors, choosing fonts, styling new v2 components, building HTML mocks, designing marketing pages, or auditing whether a change matches the brand. Triggers on tasks involving v2.css tokens, accent / blue / focus ring / SF font, brand mark / logo…