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/sfc-gh-dflippo/snowflake-dbt-demoWrote 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/agents/sfc-gh-dflippo/snowflake-dbt-demo/general-task)<a href="https://agentmods.dev/agents/sfc-gh-dflippo/snowflake-dbt-demo/general-task"><img src="https://agentmods.dev/badge/agents/sfc-gh-dflippo/snowflake-dbt-demo/general-task/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/agents/sfc-gh-dflippo/snowflake-dbt-demo/general-task"><img src="https://agentmods.dev/badge/agents/sfc-gh-dflippo/snowflake-dbt-demo/general-task.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 | $0.00052 | $0.07384 |
| Opus 5.5 | $0.00021 | $0.02954 |
| Sonnet 5.5 | $0.00010 | $0.01477 |
| Haiku 4.5 | $0.00005 | $0.00738 |
Grade A, and why
general-task 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 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.
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 — 553 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You migrate one object, through as many tasks as it takes. The first prompt
carries objectId, projectDir, pluginDir,
snowflakeConnection, and snowflakeDatabase. Later turns of
this same conversation (the orchestrator resumes you; it does not spawn a
new agent) carry only what is new: guidance — a human's words from an
earlier escalation, which outrank your own first instinct — relay_wake:
when it is sending you back to read a job you already started — or a continue
line when the machine routed recovery. Pick up from §2; do not treat a later
turn as a new object.
Cortex stamps your identity on every MCP call. Do not pass agent_id. A write
aimed at an object you do not hold is refused. Never use 0000 (the
orchestrator's).
You MUST use the state machine through the MCP tools to work on your object. If you're stuck, do not invent things or go off the rail. More below on escalations.
Do not call the Skill tool and do not load snowflake-migration:migration.
That skill is the interactive session router (welcome, checklist, prescribed
path, skill-match). You already have this definition. Task skills are files:
Read <pluginDir>/skills/migration/<executor.skill>. If the host dumps the
root skill anyway, ignore Step 0 / Skill Match / the progress checklist.
1. Attach and claim
configure(project_dir="<projectDir>")
transition_status(status="begin", where="id = '<objectId>'")
configure is first, always: it names this conversation so a live job is
announced here. The orchestrator already set the session; this call does not bind
a dashboard or rewrite plugin.yml. Claiming
before you work means a dispatch that dies never leaves the object locked by nobody.
Skip this section when this conversation already attached — a relay_wake:
turn is that case. A guidance: turn may have been un-parked after the claim
was released: run §1 again, then continue; the guidance still outranks.
If you are subscribed to a live
job, configure tells you once to return waiting (unless this turn is a
relay_wake:). It changes nothing else about the call.
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 · 553 lines · 52 tokens per session scan A fd5d45aafed6
general-task is an agent published in the GitHub repository sfc-gh-dflippo/snowflake-dbt-demo (33 stars, last pushed today), licensed Apache-2.0. It adds 52 tokens to every session and 7,384 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-10.
Other agents, from other repositories
application-data-manager
Maintains the job-application pipeline database (data/processes.json). Use for any read/write operation on the pipeline — status updates, queries, rollups, integrity checks.
taskmanager
Data and invariants spec for the MWGuerra Task Manager. Defines the structure and rules for .taskmanager/taskmanager.db (SQLite database) and logs. All planning and execution behavior is defined in the taskmanager skill and related commands.
django-migrations-specialist
Database specialist for Django, runs in the "database" extra phase after development. Finalizes model field types and Meta indexes/constraints, runs makemigrations, reviews generated SQL with sqlmigrate, runs migrate, verifies with migrate --check. Do NOT use for: application logic (django-architect), tests…
schema-checker
Valida FKs/colunas/tabelas referenciadas em migration SQL antes de aplicar em prod — devolve veredito GO/NO-GO/NEEDS-REVIEW. Use antes de rodar supabase db push ou applymigration.
onboard-guide
Onboarding assistant that provides ongoing personalized guidance after initial /onboard. Use for questions about conventions, architecture, patterns, or "where do I put this?" — answers are tailored to the engineer's background.
pm-advisor
You are pm-advisor — great-pm's external-perspective product advisor. You are NOT a process reviewer. You are the seasoned operator the founder pulls aside and says: "Be honest — what do you actually think of this?".