Conversation
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
Bundle + Package Summary:
|
| Area | Metric | Baseline | Current | Delta |
|---|---|---|---|---|
| Package | Packed tarball | 8.57 MB | 8.57 MB | +2.0 kB |
| Package | Unpacked publish size | 31.94 MB | 31.95 MB | +8.5 kB |
| Package | Installed footprint | 78.07 MB | 78.07 MB | +8.5 kB |
| Package | Published files | 3759 | 3767 | +8 |
| Package | Installed files | 7719 | 7727 | +8 |
| Package | Installed package instances | 33 | 33 | 0 |
| Package | Distinct installed package names | 32 | 32 | 0 |
| Package | Installed dependency edges | 51 | 51 | 0 |
| Package | Installed optional peer edges | 8 | 8 | 0 |
| Runtime | Unique function payloads | 2 | 2 | 0 |
| Runtime | Total function bytes | 22.06 MB | 22.08 MB | +17.4 kB |
| Runtime | Public routes | 18 | 18 | 0 |
Changed function payloads vs main (d004e6d) (2)
| Function | Status | Baseline | Current | Delta | Route changes |
|---|---|---|---|---|---|
functions/__server.func |
changed | 11.03 MB | 11.04 MB | +8.7 kB |
none |
functions/.well-known/workflow/v1/flow.func |
changed | 11.03 MB | 11.04 MB | +8.7 kB |
none |
eve init install
| Metric | Baseline | Current | Delta |
|---|---|---|---|
| Installed footprint | 116.29 MB | 116.30 MB | +8.5 kB |
| Installed packages | 98 | 98 | 0 |
| dependencies | 4 | 4 | 0 |
| devDependencies | 2 | 2 | 0 |
| Dependency package bytes | 49.38 MB | 49.39 MB | +8.5 kB |
| devDependency package bytes | 5.04 MB | 5.04 MB | 0 B ➖ |
Build Metadata
- Preset:
vercel - Nitro:
nitro@3.0.260903-beta - Output directory:
apps/fixtures/weather-agent/.vercel/output - Build metadata timestamp: 2026-09-21T01:58:21.858Z
- Route aliases: 18 public, 1 internal (19 total aliases)
- Vercel routes in config: 21
- Severity legend: 🔴 dominant/large, 🟠 notable, 🟡 watch, ⚪ small
Package Drill-Down
Package Details
- Package:
eve@0.63.0 - Package directory:
packages/eve - Tarball: 8.57 MB (
eve-0.63.0.tgz) - Unpacked payload: 31.95 MB across 3767 published files
- Installed footprint: 78.07 MB across 7727 installed files
- Installed root package: 31.30 MB
- Installed dependencies: 46.78 MB
- Installed package instances: 33
- Distinct installed package names: 32
- Installed dependency edges: 51
- Installed optional peer edges: 8
- Runtime dependencies: 2
- Peer dependencies: 6 (5 optional)
Installed footprint is measured from an isolated temporary npm install of the packed tarball.
Graph metrics read only package.json files in package directories directly beneath a node_modules boundary, including nested boundaries. Each directory is one package instance; distinct names come from those manifests. Dependency edges count each unique name in dependencies or optionalDependencies per instance; optional peer edges count peerDependencies marked optional.
Heavy installed dependencies
eve: 31.30 MB (40.1%)@rolldown/binding-linux-x64-gnu: 19.10 MB (24.5%)ai: 7.62 MB (9.8%)zod: 6.67 MB (8.5%)undici: 3.51 MB (4.5%)
Publish payload breakdown
Published file size
🔴 dist/src/compiled/shadcn-registry/index.js [#############...........] 9.70 MB 30.4%
🟠 dist/src/compiled/@photon-ai/chat-adapter-ime... [###.....................] 2.29 MB 7.2%
🟠 dist/src/compiled/@ai-sdk/code-mode/index.js [#.......................] 1.03 MB 3.2%
🟡 dist/src/compiled/@vercel/blob/index.js [#.......................] 902.9 kB 2.8%
🟡 dist/src/compiled/_chunks/workflow/signal-exi... [#.......................] 514.5 kB 1.6%
🔴 Other published files [########################] 17.51 MB 54.8%
Installed footprint breakdown
Installed package size
🔴 eve [########################] 31.30 MB 40.1%
🔴 @rolldown/binding-linux-x64-gnu [###############.........] 19.10 MB 24.5%
🔴 ai [######..................] 7.62 MB 9.8%
🔴 zod [#####...................] 6.67 MB 8.5%
🟠 undici [###.....................] 3.51 MB 4.5%
🟠 nitro [#.......................] 1.89 MB 2.4%
🔴 Other installed packages [######..................] 7.98 MB 10.2%
Runtime dependencies (2)
| Package | Range | Notes |
|---|---|---|
nitro |
3.0.260903-beta |
|
undici |
8.9.0 |
Peer dependencies (6)
| Package | Range | Notes |
|---|---|---|
@opentelemetry/api |
^1.0.0 |
optional peer |
ai |
catalog: |
|
braintrust |
^3.0.0 |
optional peer |
dd-trace |
^6.13.0 |
optional peer |
just-bash |
^3.1.0 |
optional peer |
microsandbox |
^0.5.0 |
optional peer |
eve init install drill-down
eve init install details
- Command:
eve init my-agent - Package manager:
npm - Installed footprint: 116.30 MB across 9621 installed files
- Installed packages: 98 total (92 transitive-only)
- dependencies: 4 direct packages totaling 49.39 MB
- devDependencies: 2 direct packages totaling 5.04 MB
- Other transitive package files: 61.86 MB
Installed footprint is measured from an isolated temporary eve init my-agent using the current packed eve tarball.
Heavy installed dependencies
eve: 31.30 MB (26.9%)@typescript/typescript-linux-x64: 27.95 MB (24.0%)@rolldown/binding-linux-x64-gnu: 19.10 MB (16.4%)zod: 10.29 MB (8.8%)ai: 7.62 MB (6.6%)
Installed footprint breakdown
Installed package size
🔴 eve [########################] 31.30 MB 26.9%
🔴 @typescript/typescript-linux-x64 [#####################...] 27.95 MB 24.0%
🔴 @rolldown/binding-linux-x64-gnu [###############.........] 19.10 MB 16.4%
🔴 zod [########................] 10.29 MB 8.8%
🔴 ai [######..................] 7.62 MB 6.6%
🟠 undici [###.....................] 3.51 MB 3.0%
🔴 Other installed packages [#############...........] 16.53 MB 14.2%
dependencies (4)
| Package | Range | Installed size | Share |
|---|---|---|---|
@vercel/connect |
2.2.0 |
188.7 kB | 0.2% |
ai |
^7.0.105 |
7.62 MB | 6.6% |
eve |
file:eve-0.63.0.tgz |
31.30 MB | 26.9% |
zod |
4.5.4 |
10.29 MB | 8.8% |
devDependencies (2)
| Package | Range | Installed size | Share |
|---|---|---|---|
@types/node |
24.x |
2.54 MB | 2.2% |
typescript |
7.0.2 |
2.50 MB | 2.1% |
Function Drill-Down
Payload Size Graph
Unique function payload size and share of total
🔴 functions/.well-known/workflow/v1/flow.func [########################] 11.04 MB 50.0%
🔴 functions/__server.func [########################] 11.04 MB 50.0%
Top Function Payloads
🟠 functions/.well-known/workflow/v1/flow.func • 1 public route • 11.04 MB
| Metric | Value |
|---|---|
| Public routes | /.well-known/workflow/v1/flow |
| Runtime | nodejs24.x |
| Handler | index.mjs |
| Payload | 11.04 MB |
| Function files | 11.04 MB across 117 files |
| Traced dependencies | 0 B |
| Signal | 🟠 Bundled file index.mjs is 2.89 MB (26.2%) |
🟠 🔎 Dependency Analysis
📦 Bundled files:
Bundled file size
🟠 index.mjs [##############..........] 2.89 MB 26.2%
🟡 _libs/undici.mjs [#####...................] 980.8 kB 8.9%
🟡 _chunks/sandbox.mjs [####....................] 809.9 kB 7.3%
🟡 _chunks/chatgpt-model.mjs [###.....................] 700.3 kB 6.3%
🟡 _chunks/signal-exit-B9U6kH7R.mjs [###.....................] 616.2 kB 5.6%
🔴 Other bundled files [########################] 5.04 MB 45.6%
🧾 Vercel Config
{
"handler": "index.mjs",
"launcherType": "Nodejs",
"shouldAddHelpers": false,
"supportsResponseStreaming": true,
"runtime": "nodejs24.x",
"maxDuration": "max",
"experimentalTriggers": [
{
"type": "queue/v2beta",
"topic": "__eve776561746865722d6167656e74_wkf_workflow_*",
"consumer": "default",
"retryAfterSeconds": 5,
"initialDelaySeconds": 0
}
],
"environment": {
"WORKFLOW_PRECONDITION_GUARD": "1"
}
}🟠 functions/__server.func • 17 public routes, 1 internal alias • 11.04 MB
| Metric | Value |
|---|---|
| Public routes | //.well-known/workflow/v1/webhook/[token]/eve/v1/activity/[token]/eve/v1/callback/[token]/eve/v1/connections/[name]/callback/[attemptId]/[token]/eve/v1/connections/[name]/callback/[token]/eve/v1/health/eve/v1/info/eve/v1/session/eve/v1/session/[parentSessionId]/subagents/[callId]/[childSessionId]/stream/eve/v1/session/[sessionId]/eve/v1/session/[sessionId]/cancel/eve/v1/session/[sessionId]/clear/eve/v1/session/[sessionId]/compact/eve/v1/session/[sessionId]/reset/eve/v1/session/[sessionId]/stream/eve/v1/task-input/[token] |
| Internal aliases | /__server |
| Runtime | nodejs24.x |
| Handler | index.mjs |
| Payload | 11.04 MB |
| Function files | 11.04 MB across 117 files |
| Traced dependencies | 0 B |
| Signal | 🟠 Bundled file index.mjs is 2.89 MB (26.2%) |
🟠 🔎 Dependency Analysis
📦 Bundled files:
Bundled file size
🟠 index.mjs [##############..........] 2.89 MB 26.2%
🟡 _libs/undici.mjs [#####...................] 980.8 kB 8.9%
🟡 _chunks/sandbox.mjs [####....................] 809.9 kB 7.3%
🟡 _chunks/chatgpt-model.mjs [###.....................] 700.3 kB 6.3%
🟡 _chunks/signal-exit-B9U6kH7R.mjs [###.....................] 616.2 kB 5.6%
🔴 Other bundled files [########################] 5.04 MB 45.6%
🧾 Vercel Config
{
"handler": "index.mjs",
"launcherType": "Nodejs",
"shouldAddHelpers": false,
"supportsResponseStreaming": true,
"runtime": "nodejs24.x"
}Build Timing: e2e/fixtures/agent-tools-sandbox
This is an informational timing measurement inside eve build, from preflight through publication. Output-size measurement and profile writing are excluded.
Build mode: deployable Vercel build with sandbox template prewarm included.
- Build pipeline: 3.86 s -> 3.92 s (+53.0 ms) vs
main (d004e6d). - Timing is informational: shared GitHub runners are too variable for a hard timing budget.
Detailed phase timings vs `main (d004e6d)`
| Phase | Baseline | Current | Delta |
|---|---|---|---|
extension.check |
14.1 ms | 0.4 ms | -13.7 ms |
project.resolve |
0.4 ms | 0.3 ms | -0.1 ms |
workspace.create |
0.6 ms | 0.5 ms | -0.1 ms |
host.prepare |
533.9 ms | 599.6 ms | +65.7 ms |
vercel.service-prefix.resolve |
2.1 ms | 2.5 ms | +0.4 ms |
nitro.create |
294.2 ms | 288.0 ms | -6.2 ms |
sandbox.prewarm |
294.3 ms | 291.6 ms | -2.7 ms |
nitro.cache.prepare |
0.2 ms | 0.3 ms | +0.1 ms |
nitro.prepare |
0.7 ms | 0.8 ms | +0.1 ms |
nitro.public-assets |
0.7 ms | 0.7 ms | 0.0 ms |
nitro.prerender |
0.4 ms | 0.5 ms | +0.1 ms |
nitro.bundle |
2.66 s | 2.68 s | +12.0 ms |
nitro.cache.write |
0.4 ms | 0.3 ms | -0.1 ms |
vercel.workflow-function.materialize |
49.9 ms | 48.1 ms | -1.8 ms |
agent-summary.emit |
0.6 ms | 0.6 ms | 0.0 ms |
nitro.close |
0.1 ms | 0.1 ms | 0.0 ms |
output.publish |
3.5 ms | 3.1 ms | -0.4 ms |
workspace.remove |
2.3 ms | 2.1 ms | -0.2 ms |
![e0-gh-vercel-connect[bot]](https://avatars.githubusercontent.com/b/10660?s=60&v=4){kind=small}
| readonly instructions?: string; | ||
| readonly message: string; | ||
| /** Evaluation model instance or ID. Defaults to TypeSafe Jev. */ | ||
| readonly model?: Parameters<typeof evaluate>[0]["model"]; |
There was a problem hiding this comment.
P1: Do not accept live model instances across this step boundary. auto() is itself a "use step" function, so this options object is serialized when callers use it from the documented workflow body. Evaluation-model instances contain executable doEvaluate behavior and cannot cross that durable boundary, so the documented model: providerModel path fails before routing. The unit test calls auto() directly, and the E2E fixture has one candidate and returns before evaluation, so neither exercises this path. Please restrict and validate this as a non-empty model ID like AgentRouterOptions, then remove the instance claim from the docs, or redesign the API so the model is constructed inside the registered step from serializable configuration.
| throw new Error("agentRouter auto requires a non-empty model ID when provided."); | ||
| } | ||
| const criteria = Object.fromEntries( | ||
| Object.entries(agents).flatMap(([name, description]) => { |
There was a problem hiding this comment.
P2: Validate candidate names at this public boundary. auto() now accepts a user-authored map, but only its descriptions are validated. For example, { " ": "Research" } reaches the sole-candidate fast path and returns " ", while ctx.agent() rejects any target whose trimmed name is empty. Please reject empty or whitespace-only keys while building criteria, before either the fast path or evaluation.
There was a problem hiding this comment.
While I don't think this is likely I think you do want to check this and figure out what backtracking could look like.
| export default agentRouter(); | ||
| ``` | ||
|
|
||
| Pass `model` or `instructions` to customize the provided router. Because factory options become durable workflow configuration, `model` must be a serializable model ID; use `auto()` directly when you need an evaluation model instance: |
There was a problem hiding this comment.
P3: Describe the configured evaluator. Now that agentRouter({ model }) supports a custom model, the behavior text below is inaccurate when it says the router always sends the request to JEV. Please describe this as the configured evaluation model, with JEV as the default.
![vercel[bot]](https://avatars.githubusercontent.com/in/8329?s=60&v=4){kind=small}
| throw new Error("agentRouter auto requires a non-empty model ID when provided."); | ||
| } | ||
| const criteria = Object.fromEntries( | ||
| Object.entries(agents).flatMap(([name, description]) => { |
There was a problem hiding this comment.
normalizeAgents() validates agent descriptions but never validates the agent name keys, so an empty/whitespace-only key survives and is returned as a route target by the public auto() helper, deferring the failure to a confusing downstream error.
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
ctgowrie
changed the title
…o-helper Signed-off-by: Casey Gowrie <ctgowrie@gmail.com> # Conflicts: # packages/eve/extension-contracts/reports/tool/v53.json # packages/eve/src/compiler/extension-compatibility.ts
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
Signed-off-by: Casey Gowrie <ctgowrie@gmail.com>
|
Do not think this should be part of the public API |
This branch was successfully deployed
3 participants
Summary
Custom workflow routers currently have to repeat
agentRouter()'s JEV selection logic. This exports reusableauto()fromeve/tools/agent-router/autofor authored routing steps, including steps that import a live evaluation model, and addsagentRouter({ select })for plugging an authored durable selector into the provided router. Withoutselect,agentRouter()keeps its internal default step and optional serializable model-ID/instruction overrides; its exact zero-argument API remains supported.The workflow guide presents three choices in order: write the full
evaluate()step, callauto()inside an authored step, or useagentRouter()with its default or a custom selector. Custom selector names are validated against the described candidate map before invocation.Validation
The full eve unit suite passed 8,987 tests with 1 skipped. Deterministic e2e coverage passed 5/5 gates with two candidates:
agentRouter({ select })reconstructs the authored selector step, that step imports a live evaluation model and callsauto(), and the selected root copy is invoked. eve and fixture typecheck, docs, lint, invariants, and retained extension-contract checks also pass.Checklist
CONTRIBUTING.mdevepackagegit commit --signoff)Diff size
Docs — 3 files ·
+50 / -3Documents the full custom pattern, reusable
auto()logic, and provided router with default or authored selector as three distinct choices, plus release-facing behavior.Implementation — 11 files ·
+216 / -37Extracts reusable selection logic, carries authored selectors by stable step ID, reconstructs selector proxies in the workflow body, validates options/results, keeps model-auth code behind the step boundary, and exports the new subpath and types.
Tests — 10 files ·
+289 / -3Covers defaults, live and string evaluation models, filtering, validation, selector transport and result checks, deterministic workflow bundling/execution, and retained tool capability compatibility.