Execution profile — dbt platform (Studio)

Mechanics for running upgrading-dbt inside a Studio develop session. The rules, the phase order, and when to do each of these live in SKILL.md; this file only says how. If you have a shell, you want exec-local.md instead.

What is different here

There is no shell. The skill's scripts/tools.py cannot run, and neither can anything else you might be tempted to shell out to. Every step the script does locally is done with a Studio tool below.

Three consequences worth stating plainly:

• dbt_command accepts exactly two binaries: dbt and dbt-autofix. Nothing else is on the allowlist — notably not dbt-migrate-1x (see autofix below), which is why this migration's deterministic fixes are applied by hand here.
• You write the artifacts yourself with edit_file. No script validates them, so the schemas in SKILL.md are the contract — never invent a field, a status value, or a phase id.
• There is no write_file. edit_file creates the file when the path does not exist, so it is also how you create an artifact for the first time.

Never install anything, and treat environment_change issues as advisory edits only — same rule as local, but here there is not even a mechanism to break it.

State layout — one file, yours to maintain

All migration state lives in a single file, target/dbt_migration.json:

{
  "version": 1,
  "updated_at": "2026-08-13T09:15:00+00:00",
  "steps": [
    { "id": "preflight", "label": "Git preflight",
      "status": "complete", "note": "On branch upgrade/dbt-1.12, clean tree" },
    { "id": "detect", "label": "Detection sweep",
      "status": "in_progress", "note": "12 of 41 issues checked" }
  ],
  "issues": {
    "1_7_003": {
      "automation_type": "agentic",
      "out_of_repo_risk": false,
      "environment_change": false,
      "status": "fixed",
      "files_changed": ["models/marts/customers.sql"],
      "notes": "renamed + rewrote ref"
    }
  },
  "outcome": {
    "outcome": "completed_partial",
    "reason": null,
    "gate": { "parse": "passed", "compile": "passed", "test": "failed", "build": "not_run" },
    "repair_attempts": 2,
    "files_changed_count": 12,
    "flags_pinned_count": 2,
    "manual_actions_count": 3,
    "jobs_to_flip_count": 1
  }
}

steps and issues carry exactly the records SKILL.md's Migration state section defines — same phase ids, same status values, same fields. Only the packaging differs: locally these are two script-written files, here they are two keys in one file you write with edit_file.

outcome is absent until Step 7 exits. Write gate and repair_attempts then, and the remaining fields in Step 9 (or just before stopping, if the run ends early). SKILL.md's Run outcome defines its fields and values. The gate_result and outcome telemetry events are filled from it.

One file, because you are writing it by hand and nothing checks your work. Two files can disagree — a phase marked complete while its issues still read pending — and reconciling them costs edits you would rather spend on the migration. This file is also what you read at the end to write the report, so keeping it whole and current is what makes Step 9 possible.

Refresh updated_at on every write. Never rewrite the document wholesale from memory: read it, change the one record you mean to change, write it back.

Tools you have

You do not trigger jobs. trigger_job_run is not part of this skill. A job run executes on the deployment environment's own credentials and its own target schema, which is very often production — the migration has no business writing there, and a schema_override is a flag you can forget. Verification stays in this session, on this session's development credentials. See verify-commands.

$PROJECT = the project's root. $ADAPTER = the adapter type. $FROM = the starting version. <id> = an issue_id from the bundle.

The adapter needs no lookup — it is already in your system prompt as dialect.

Operations

status-init

edit_file on target/dbt_migration.json, creating it with all ten phases present and pending, in execution order, and an empty issues map. Use the exact phase ids and status values from SKILL.md's Migration state section.

status-set

edit_file the same file, changing that one phase's status and note, and refreshing updated_at. Change one row at a time — do not rewrite the whole document from memory, or you will drop state you have already recorded.

preflight

git status and git branches. If the session is on main/master or the tree is dirty, request_user_input before going further. Create the migration branch with git checkout and create_if_missing. If the run stops here, send gave_up first: reason preflight_blocked if git stayed unsafe, user_aborted if the user said not to continue.

Studio also refuses commits to protected branches, so this is guarded twice — but do the check anyway; the point is to tell the user before doing work, not to be caught later.

load-bundle

load_skill_resource_file → references/kb_<FROM>_<ADAPTER>.json

e.g. references/kb_1_5_snowflake.json. Versions are dotless (1.5 → 1_5) and the adapter is core when there is none.

This is the whole corpus for this migration — self-contained, with every issue's action, automation_type, and context.detection / context.fixing. The kb/ YAML corpus these are compiled from does not ship in the package; do not try to read it.

init-results

edit_file the same target/dbt_migration.json, filling its issues map with one record per issue in the bundle at status: "pending", copying automation_type, out_of_repo_risk and environment_change straight from the bundle.

If records are already there, keep the statuses they carry — that is what makes a run resumable.

set-status

edit_file that same file, updating one issue's status, files_changed, and notes under issues. One record at a time, for the same reason as status-set.

list-issues

read_file on target/dbt_migration.json and filter the issues map yourself. There is no query tool; the file is small enough to read whole.

autofix

dbt-autofix migrate-1x is Not available here.

So there is no separate autofix pass in this profile. Treat every detected issue with automation_type: deterministic exactly like an agentic one in Step 5: apply the fix yourself per context.fixing, git diff to see what changed, then set-status fixed — never handled-by-autofix, which claims a tool did it and nothing did here. Skip Step 4 as its own pass (there is nothing for it to do); fold its issues into Step 5's work instead, and status-set autofix = complete immediately with a note saying why ("no dbt-migrate-1x tool in Studio; N deterministic issues folded into Step 5").

set-flag

edit_file on dbt_project.yml, adding the flag named in that issue's behavior_flag.name under flags: set to false.

Add the key if flags: already exists rather than replacing the block, and pin only flags for behaviors detection actually found.

parse

dbt_command with dbt parse. Poll with dbt_command_status. When the 5-attempt repair cap in SKILL.md Step 7 is reached, send gave_up (scope gate, reason parse_cap_reached) before you revert anything.

This session already runs on the target release track. The platform sets your personal version override to the target, and Studio starts or restarts the development session before handing off to you. Assume the override is applied. You do not need to check the running version, and you must not try to change it.

What the gate is: the cheap check. Run it before verify-commands, because it catches the small mistakes — a bad ref, malformed YAML, a config that moved — in seconds instead of in a dbt build that spends warehouse compute to tell you the same thing. It proves the project parses. It cannot prove behaviour; that is verify-commands.

jobs-file

You create this file here. No extension ran before you, so unlike the local profile there is nothing on disk to read — you discover the jobs yourself, then record your verdicts in it.

1. list_jobs for this project. project_id is in your context; the account is supplied by the tool and is never yours to pass.
2. Keep only the jobs on a legacy version — below 1.8. Use each job's own pinned version if it has one, otherwise its environment's.
3. For each kept job, get_job_details → its execute_steps.
4. edit_file on migration_jobs.json at the project root (not under target/), creating it with one entry per job and one step entry per command, every step status: "pending" and updated: null.
5. Then edit_file the same file as you work through the issues, setting each step to ok / needs_change / manual.

Set "source": "platform". Do not list an account's other projects, and never edit or run a job.

If list_jobs or get_job_details fails, do not stop the migration. Write the file with whatever jobs you could read (an empty jobs list if none), carry on, and list "review job commands by hand: the job list could not be read" in the report as a manual action.

The schema is fixed and shared with the VS Code extension, which writes the same file deterministically on the local path: see Job commands and match it exactly. Nothing validates it for you here, so that section is the contract — do not add fields, rename them, or invent status values.

Write it in Step 2, before detection, not when the first job-command issue turns up. Later steps record verdicts against it, so it has to exist whether or not any out-of-repo issue was found.

This file is for tracking and reporting only. verify-commands does not read it — that gate runs a fixed set of commands, listed below. What this file is for is telling the customer which of their job commands they have to change after merging. Nothing in this skill edits a job, and nothing in this skill runs one.

revert

git revert with a files list. That undoes those uncommitted changes, which is what git restore does locally. There is no stash.

report

read_file on target/dbt_migration.json and edit_file to write migration_report.md from its issues map, grouped by outcome, then show it in chat. Locally a script renders this; here you render it yourself, which is the reason the state file has to have been kept accurate all the way through. Then write the state file's outcome object and send the outcome event, both before status-set → report = complete.

Cover what changed, which behavior flags were pinned and why, anything left manual-required or failed, and — for this environment specifically — every environment and job the user still has to flip, listed together. A partial flip leaves the project split across release tracks.

For job commands, give the count and link migration_jobs.json rather than restating them; that file is the actionable list and a prose copy will drift from it. Flipping a job's version and fixing its commands are two different jobs of work — say both.

ask

request_user_input. Always status-set the current phase to waiting_input before asking, with a note saying what you asked, and set it back to in_progress the moment they answer.

verify-commands — the command checks

Platform only. These are the checks above parse in SKILL.md's Step 7. dbt parse cannot catch behavior-only changes — connector swaps, quoting, timeout defaults, a changed materialization default — and those are exactly what breaks after a version bump. Commands that compile and build can.

Same tool as the parse check: dbt_command, polled with dbt_command_status.

The commands are fixed. Run these four, in this order:

1. dbt parse — check 1, already run. Do not repeat it here.
2. dbt compile
3. dbt test
4. dbt build

Stop at the first red. Do not substitute, add or reorder them, and do not take commands from migration_jobs.json.

dbt deps first if the project needs its packages. That is setup, not verification, and it is the one extra command you may run.

Never pass --target, --profile, or --profiles-dir. The session already runs on the signed-in user's own development credential and schema, which is where these commands are meant to build.

The loop. ask for approval once, up front, saying that dbt build will materialize the project into the development schema. Then for each command in order:

1. dbt_command with the command, no extra selectors.
2. Poll with dbt_command_status. Green → next command.
3. Red → read the node-level errors in the status output, attribute the failure to an issue, go back to Step 5 or 6, then re-run Step 7 from the parse check. Record what you changed in the issue's notes. For a warehouse failure that no issue explains (SKILL.md Step 7), send gave_up (scope gate, reason warehouse_error).

When all four are green, go on to Step 8. Whenever the gate exits, green or not, send gate_result once.

Guardrails:

• One command at a time. Do not fan out; dbt_command is per-command anyway and concurrent builds into one schema will collide.
• Max 3 trips round the loop, then stop. Unlike a job trigger there is no per-run approval bounding this, so the cap is what bounds it. Say you hit the cap rather than quietly stopping. Send gave_up (scope gate, reason command_cap_reached) before you stop.
• Never on main/master — always the migration branch, same as everything else in this skill.

If you cannot run it — the user declines, the session has no warehouse connection, the project's schema generation is not safe to build into — that is a normal outcome. Report verification as incomplete, retain any parse result, and make sure the report names every command left unverified so the user knows exactly what was and was not proven. Send gave_up (scope gate) first, with reason no_warehouse_connection when there is no connection and verification_declined otherwise. Those unrun commands are declined in gate_result.

Telemetry (platform only)

record_migration_event is a server tool. It records how the run went so the dbt platform can measure migrations. Call it only if it is in your tool list. It is telemetry only:

• It never blocks the migration or changes any decision in it.
• If a call errors, carry on and do not retry.

Every call passes kind plus that kind's fields:

error_category ∈ syntax · deprecated_config · macro · adapter · warehouse_auth · warehouse_timeout · test_failure · other.

gave_up reason, one per stop path:

outcome values are defined in SKILL.md, Run outcome.