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 obto-inc/platform --skill obto-databasegit clone --depth 1 https://github.com/obto-inc/platformWrote 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/obto-inc/platform/obto-database)<a href="https://agentmods.dev/skills/obto-inc/platform/obto-database"><img src="https://agentmods.dev/badge/skills/obto-inc/platform/obto-database/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/obto-inc/platform/obto-database"><img src="https://agentmods.dev/badge/skills/obto-inc/platform/obto-database.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 | $0.00176 | $0.03349 |
| Opus 5.5 | $0.00070 | $0.01340 |
| Sonnet 5.5 | $0.00035 | $0.00670 |
| Haiku 4.5 | $0.00018 | $0.00335 |
Grade B, and why
obto-database scanned grade B 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 12d 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.
Subtle steeringmediumPrompt injection
Instructions that bias recommendations or shape behaviour without the user noticing.
6. **Follow it to the end** with `obto_pg_status` (state, per-unit `run`, `ledgerSeq`). Done means: state `applied`, a `ledgerSeq`, each unit's `expect` true when you run it through `obto_pg_query`, and `obto_pg_drift` o How it starts
The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OBTO App Database (Postgres)
Some OBTO apps own a Postgres database. The platform holds the connection; you reach the database only through the obto_pg_* tools, and every call carries appName (the app that owns the database) and your domain (from obto_whoami). config.unknown_app means the app has no database registered on OBTO — ask the OBTO operator; do not look for another way in.
| Tool | Use it to | Who |
|---|---|---|
obto_pg_query |
read data and the catalog (one read-only statement) | you |
obto_pg_check |
lint a schema change against the live catalog; stores nothing | you |
obto_pg_propose |
prepare: lint + rehearse on the real database inside BEGIN … ROLLBACK, then store the change |
app owner, OBTO |
obto_pg_status / obto_pg_list |
follow one change / list changes, newest first | you |
obto_pg_retry / obto_pg_cancel |
re-lint and re-rehearse a blocked change / stop one that has not run | app owner, OBTO |
obto_pg_drift / obto_pg_ledger |
compare the live schema with the ledger / read the change history | you |
obto_pg_apply |
run a prepared change on production, after the owner's yes in chat | app owner, OBTO |
Only the app's owner and OBTO can change the app's database. The owner is a user whose home domain is the app's domain and who has the super role there; a user who reaches the domain through a cross-tenant switch never is. The owner's AI checks a change, prepares it (obto_pg_propose) and applies it (obto_pg_apply) after the owner says yes in chat. Anyone else reads and checks only: every write answers auth.owner_only ("Only the app's owner can change this database. Ask them."). auth.owner_changes_off means OBTO has paused owner changes: ask OBTO. Always OBTO: a baseline and a change that carries prismaRefreshImage (OBTO builds and rolls that image). These are expected answers, not bugs.
If these tools are missing from your tool list, the connection predates the release that added them — see "After an OBTO update" in obto-troubleshooting.
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.
- 12d ago Changed · -6 lines · -5 tokens per session d6ce71dcba3f
- 16d ago Changed · +7 lines · +14 tokens per session cc3763c4fbb5
- 17d ago First seen · 124 lines · 167 tokens per session scan B 567cb200986a
obto-database is a skill published in the GitHub repository obto-inc/platform (1 stars, last pushed 12d ago), licensed Apache-2.0. It adds 176 tokens to every session and 3,349 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it B with 1 finding (subtle steering). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-20.
obto-database compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| obto-databaseobto-inc/platform | Skill | 1 | 12d ago | Apache-2.0 | B | 176 | Claude Code, Codex |
| amazon aurora dsqlawslabs/mcp | Skill | 9.9k | yesterday | Apache-2.0 | A | 46 | Claude Code |
| temps-add-migration-and-entitygotempsh/temps | Skill | 903 | yesterday | Apache-2.0 | A | 84 | Claude Code, Codex |
| db-sculptorEliasOulkadi/shokunin | Skill | 112 | 1mo ago | MIT | A | 123 | Claude Code |
| neon-postgresEliasOulkadi/shokunin | Skill | 112 | 1mo ago | MIT | A | 76 | Claude Code, Codex |
| db-adminEliasOulkadi/shokunin | Skill | 112 | 1mo ago | MIT | A | 122 | Claude Code |
| vector-backend-operationsKnuckles-Team/vector-mcp | Skill | 16 | yesterday | MIT | A | 113 | Claude Code, Codex |