Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Migrate the schema / data on a live app · get-convex | DecimalAI

get-convex/convex-migrate

NewSkillScoreMaintained

Migrate schema + backfill data on a deployed Convex app using @convex-dev/migrations.

Databy get-convex

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/get-convex-convex-migrate/SKILL.md

decimalai skills pull get-convex-convex-migrate --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/get-convex-convex-migrate/SKILL.md https://app.decimal.ai/s/get-convex-convex-migrate/SKILL.md

Raw file: /s/get-convex-convex-migrate/SKILL.md ↗ · machine-readable summary at /s/get-convex-convex-migrate.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 25h ago

Yes — +50% lift-17% tokens0% turns68% of whole cases pass with skill (15/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-02	✗→✓	▲ Improved	-22%	0%
case-04	✗→✓	▲ Improved	-18%	0%
case-05	✗→✓	▲ Improved	-23%	0%
case-08	✗→✓	▲ Improved	-19%	0%
case-10	✗→✓	▲ Improved	-23%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/7/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

get-convex / agent-skills

skills/convex-migrate/SKILL.md

View on GitHub ↗

Imported 8/7/2026Version v1

Skill contents

SKILL.md

Migrate the schema / data on a live app

Change a deployed schema without breaking existing data: stage the schema change, install @convex-dev/migrations, write a backfill that makes old rows valid, run it, and verify before tightening the validator.

Workflow

Make the new field optional first (so deploy doesn't reject existing rows).
Install @convex-dev/migrations; write a migration that backfills/transforms existing rows.
Run the migration; verify all rows are valid.
Tighten the validator (make the field required) once the backfill is complete.

Rules

Never tighten a validator before the backfill completes — it rejects existing rows and breaks the live app.
Add new fields as optional first, migrate, then require.
Verify row counts before and after.

Want to see if get-convex/convex-migrate would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/28/2026, 10:36 UTC · 22 cases

Lift

+50%

68% vs 18%

Agent turns

0%

1.0 vs 1.0

Tokens

-17%

2,081.5 vs 2,497.18

fail → pass · 11pass → pass · 4fail → fail · 7

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I need to add a new required `tenantId` field to my `documents` table in a live Convex app. Since there are already thousands of active records without this field, how should I handle the schema change and data update so production traffic isn't disrupted?	fail→fail	15,427	22,617	+47%	1	1	0%	2,991	2,570	-14%	0	0	—
case-02 We are updating our Convex database schema to make an existing optional user property mandatory. Could you provide a complete step-by-step guide and code pattern to safely backfill the legacy rows and update our schema validator?	fail→pass	22,059	11,817	-46%	1	1	0%	3,062	2,387	-22%	0	0	—
case-03 I'm refactoring a table in my deployed Convex backend by splitting a `name` field into `firstName` and `lastName`. What is the proper procedure for transforming the existing documents and adjusting the schema definitions without causing validation failures?	fail→fail	14,721	15,050	+2%	1	1	0%	2,723	2,449	-10%	0	0	—
case-04 I need to add a required `organizationId` string field to an active `projects` table in my production Convex backend. Walk through the exact schema definition changes and migration steps required to do this safely without failing deployments.	fail→pass	13,926	9,747	-30%	1	1	0%	2,574	2,111	-18%	0	0	—
case-05 In a Convex project, we want to rename the field `user_email` to `email` on the live `users` table. Provide the code architecture and step-by-step migration sequence so no queries break during rollout.	fail→pass	21,487	14,918	-31%	1	1	0%	3,273	2,515	-23%	0	0	—
case-06 Our production Convex `tasks` table needs a new `status` string field that defaults to `'pending'` for all existing rows. How should we configure the Convex schema and execute the data backfill safely?	fail→fail	19,624	12,887	-34%	1	1	0%	2,920	2,165	-26%	0	0	—
case-07 We are storing Unix timestamps as strings in our Convex `events` table and want to convert them to numbers. Outline the technical steps to transform the field in a live system without causing validation errors.	fail→fail	16,597	10,409	-37%	1	1	0%	2,884	2,196	-24%	0	0	—
case-08 Our Convex schema currently defines `role: v.string()` on the `accounts` table, but we need to migrate it to `roles: v.array(v.string())`. Explain how to transition the schema and existing documents in Convex.	fail→pass	15,104	13,743	-9%	1	1	0%	2,914	2,369	-19%	0	0	—
case-09 We need to replace a top-level `address` string field in Convex with a structured `address` object containing `street` and `city`. Describe the safe rollout sequence and tooling to handle this Convex schema migration.	fail→fail	17,084	15,886	-7%	1	1	0%	3,069	2,609	-15%	0	0	—
case-10 How do I safely introduce a required boolean `isArchived` flag (defaulting to `false`) to a Convex table containing over 100,000 live rows?	fail→pass	16,479	11,826	-28%	1	1	0%	3,132	2,403	-23%	0	0	—
case-11 We have an existing optional field `profilePictureUrl` in Convex, and we now want to make it mandatory with a default fallback URL for users missing it. What is the correct step-by-step migration procedure?	fail→pass	10,293	10,113	-2%	1	1	0%	1,879	2,178	+16%	0	0	—
case-12 We are updating integer prices to floating point numbers in our active Convex `orders` database table. Describe how to perform this migration safely and verify success.	fail→fail	23,299	16,810	-28%	1	1	0%	3,403	2,626	-23%	0	0	—
case-13 We are deprecating a field `legacyCode` from a live Convex database. What is the proper process for removing a field from a Convex schema without breaking live deployment or leaving orphan data?	fail→pass	15,643	10,671	-32%	1	1	0%	2,571	2,210	-14%	0	0	—
case-14 A developer tried adding `apiKey: v.string()` directly to a Convex schema file on a live app, but deployment failed due to existing documents missing the field. What exact sequence of changes should they follow to fix this?	fail→pass	10,890	8,526	-22%	1	1	0%	1,501	1,437	-4%	0	0	—
case-15 We want to ensure every record in our Convex `posts` table has a `tags` field initialized to an empty array `[]`. Provide the schema definition and migration code structure for this operation.	fail→fail	15,306	9,024	-41%	1	1	0%	2,132	2,016	-5%	0	0	—
case-16 In our Convex database, we want to restrict an existing string `category` field to a union of literals `v.union(v.literal('a'), v.literal('b'))`. How should we prepare existing data and update the schema validator?	fail→pass	16,441	9,942	-40%	1	1	0%	2,394	2,007	-16%	0	0	—
case-17 We are adding a calculated `fullName` field derived from `firstName` and `lastName` across existing documents in a Convex application. Detail the process for adding this field and tightening schema validation.	fail→pass	16,319	10,103	-38%	1	1	0%	3,048	2,209	-28%	0	0	—
case-18 We want to add a nested `metadata` object containing mandatory `createdAt` and `updatedAt` properties to all rows in an active Convex table. What is the recommended rollout procedure?	fail→pass	16,853	10,047	-40%	1	1	0%	2,496	2,200	-12%	0	0	—
case-19 Explain the order of operations between updating Convex schema definitions with `v.optional()` vs `v.string()` and running data migration scripts during live updates.	pass→pass	18,243	12,055	-34%	1	1	0%	2,559	1,890	-26%	0	0	—
case-20 I am building a brand new Convex app from scratch with no existing production data. How do I define a schema in `convex/schema.ts` for a `users` table with required `email` and `name` fields?	pass→pass	5,044	6,386	+27%	1	1	0%	704	981	+39%	0	0	—
case-21 I need to write a standard Convex mutation in `convex/tasks.ts` that inserts a single new task document with `title` and `completed` fields when invoked by a frontend user. Provide the mutation code.	pass→pass	8,810	3,862	-56%	1	1	0%	1,281	936	-27%	0	0	—
case-22 How do I add a database index on the `userId` field of an existing `orders` table in my Convex `schema.ts` file to optimize query performance?	pass→pass	9,908	5,970	-40%	1	1	0%	1,428	1,329	-7%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +50 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 18.2%
Pass rate with the skill: 68.2%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/28/2026, 10:36 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-02

Prompt

We are updating our Convex database schema to make an existing optional user property mandatory. Could you provide a complete step-by-step guide and code pattern to safely backfill the legacy rows and update our schema validator?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested yesterday

Other measured skills in the registry, with their headline benchmark lift.