Implemented charge windows instead of just picking random low price s… by fredli74 · Pull Request #951 · fredli74/smartcharge-dev

fredli74 · 2026-02-06T01:35:11Z

No description provided.

…egments Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 · 2026-02-06T01:38:48Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f4aff26557

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

server/logic.ts

Copilot

Pull request overview

This pull request implements a sophisticated charge window scheduling algorithm that replaces the simpler approach of just picking random low-price slots. The new system intelligently schedules charge windows by considering price data, time constraints, and multiple competing charge intents (manual, trip, routine, fill) to optimize both cost and charging behavior.

Changes:

Replaced the old GeneratePlan function with a new "soft intents" model that collects all charge requirements first, then schedules them optimally across available price windows
Implemented a greedy sliding-window algorithm to find optimal charge windows that minimize cost while respecting constraints like minimum window duration (to avoid fragmentation in cold weather) and maximum price thresholds
Added support for splitting charge sessions across multiple windows when cost savings justify it, with configurable thresholds for when to split vs. consolidate charging

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

server/logic.ts

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

server/logic.ts

fredli74 · 2026-02-06T19:11:19Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 289a128246

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

server/logic.ts

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

server/logic.ts

app/src/app.vue

server/logic.ts

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

… quantumMs Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

…ehicle not waking up to charge, cause it though the level was ok already Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

app/src/app.vue

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

…tting logic.

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 · 2026-02-18T09:34:18Z

@codex review

Copilot

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 6 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-18T09:37:41Z

server/logic.ts

+        // DP over (startSlot, neededSteps). Gap penalty is modeled as extra score cost, not extra time.
+        const dpTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));
+        const dpWindows = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));
+        const choiceTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(null));


The dynamic programming algorithm creates three large 2D arrays (dpTable, dpWindows, choiceTable) with dimensions (numSlots + 1) × (timeNeededSteps + 1). With many price slots and long charging durations, this could consume significant memory. For example, 100 slots and 1000 steps would create arrays with 100,000 elements each. Consider adding bounds checking or logging warnings when the arrays become very large.

Suggested change

// DP over (startSlot, neededSteps). Gap penalty is modeled as extra score cost, not extra time.

const dpTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));

const dpWindows = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));

const choiceTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(null));

// Estimate DP memory usage and warn if very large.

const dpRows = numSlots + 1;

const dpCols = timeNeededSteps + 1;

const dpCellsPerTable = dpRows * dpCols;

const dpTablesCount = 3;

const totalDpCells = dpCellsPerTable * dpTablesCount;

const LARGE_DP_THRESHOLD = 500_000;

if (totalDpCells > LARGE_DP_THRESHOLD) {

vehicleLog(

LogLevel.Warn,

vehicle.vehicle_uuid,

`scheduleWindows(${scheduleTag}): large DP allocation: ` +

`slots=${numSlots} steps=${timeNeededSteps} cellsPerTable=${dpCellsPerTable} totalCells=${totalDpCells}`

);

}

// DP over (startSlot, neededSteps). Gap penalty is modeled as extra score cost, not extra time.

const dpTable = Array.from({ length: dpRows }, () => new Array(dpCols).fill(Infinity));

const dpWindows = Array.from({ length: dpRows }, () => new Array(dpCols).fill(Infinity));

const choiceTable = Array.from({ length: dpRows }, () => new Array(dpCols).fill(null));

Copilot · 2026-02-18T09:37:42Z

providers/tesla/tesla-agent.ts

      // Convert to numeric charge plans with start and stop times rounded to 15 minutes
      .map((p): NumericChargePlan => {
-        if (wantedSoc === undefined && p.level) {
+        if (p.level !== undefined && (wantedSoc === undefined || wantedSoc < p.level)) {


The condition has been changed from checking if wantedSoc === undefined to checking if p.level !== undefined && (wantedSoc === undefined || wantedSoc < p.level). This changes the logic to update wantedSoc to the maximum level from all charge plans, not just the first one encountered. While this may be intentional to find the highest charge level needed, it could cause unexpected behavior if multiple plans specify different levels and the intent was to use a specific one. Verify this behavior is correct for your use case.

Suggested change

if (p.level !== undefined && (wantedSoc === undefined || wantedSoc < p.level)) {

if (wantedSoc === undefined && p.level !== undefined) {

Copilot · 2026-02-18T09:37:42Z

server/gql/vehicle-resolver.ts

+    vehicleLog(LogLevel.Debug, input.id, `updateVehicle: ${JSON.stringify(input)}`);
    await context.db.getVehicle(accountFilter(context.accountUUID), input.id);


The vehicleLog is being called with input.id before the vehicle is retrieved from the database. If the vehicle doesn't exist or the user doesn't have access, this will log with an incorrect vehicle UUID. The log should be moved after line 143 where the vehicle is retrieved, and use vehicle.vehicle_uuid instead.

Suggested change

vehicleLog(LogLevel.Debug, input.id, `updateVehicle: ${JSON.stringify(input)}`);

await context.db.getVehicle(accountFilter(context.accountUUID), input.id);

const vehicle = await context.db.getVehicle(accountFilter(context.accountUUID), input.id);

vehicleLog(LogLevel.Debug, vehicle.vehicle_uuid, `updateVehicle: ${JSON.stringify(input)}`);

Copilot · 2026-02-18T09:37:42Z

server/gql/vehicle-resolver.ts

    // verify vehicle ownage
-    log(LogLevel.Debug, `updateSchedule: ${JSON.stringify(id)} ${JSON.stringify(vehicleID)}`);
    const vehicle = await context.db.getVehicle(accountFilter(context.accountUUID), vehicleID);
+    vehicleLog(LogLevel.Debug, vehicle.vehicle_uuid, `updateSchedule: ${JSON.stringify(id)} ${JSON.stringify(vehicleID)}`);


The vehicleLog is being called with input.id before the vehicle is retrieved from the database. If the vehicle doesn't exist or the user doesn't have access, this will log with an incorrect vehicle UUID. The log should be moved after line 216 where the vehicle is retrieved, and use vehicle.vehicle_uuid instead.

Copilot · 2026-02-18T09:37:42Z

providers/tesla/tesla-api.ts

+      log(LogLevel.Trace, `TeslaAPI.renewToken(${redactSecret(refresh_token)}) response: ${JSON.stringify({ expires_in: authResponse?.expires_in })}`);
      return this.parseTokenResponse(authResponse);
    } catch (e) {
      console.debug(`TeslaAPI.renewToken(${refresh_token}) error: ${e}`);


The error handler logs the unredacted refresh_token in the error message. This could expose sensitive credentials in logs. Consider using redactSecret(refresh_token) here as well, consistent with the success path logging.

Copilot · 2026-02-18T09:37:43Z

server/logic.ts

+        setSmartStatusFromIntent(allocations[0], "scheduled");
+        const sorted = windows.slice().sort((a, b) => a.start - b.start);
+        let i = 0;
+        let remaining = allocations.length > 0 ? allocations[0].durationMs : 0;


The applyWindows function accesses allocations[0] without checking if the array is empty. While line 1028 calls setSmartStatusFromIntent(allocations[0], "scheduled") which safely handles undefined, if the function is ever called with an empty allocations array from other contexts, it could cause issues. Consider adding a guard check at the beginning of the function to return early if allocations is empty.

Suggested change

setSmartStatusFromIntent(allocations[0], "scheduled");

const sorted = windows.slice().sort((a, b) => a.start - b.start);

let i = 0;

let remaining = allocations.length > 0 ? allocations[0].durationMs : 0;

if (!allocations.length) {

return;

}

setSmartStatusFromIntent(allocations[0], "scheduled");

const sorted = windows.slice().sort((a, b) => a.start - b.start);

let i = 0;

let remaining = allocations[0].durationMs;

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8ec11594a7

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-02-18T09:41:00Z

server/logic.ts

+          if (i.beforeTs !== undefined) intentDeadline = Math.min(intentDeadline, i.beforeTs);
+          if (i.level > intentMaxLevel) intentMaxLevel = i.level;


Keep each soft intent deadline when building charge windows

In scheduleSoftIntents, the planner collapses all soft intents into intentDeadline = min(beforeTs) and intentMaxLevel = max(level), then schedules that max level against the earliest deadline. This couples unrelated targets and can force high future targets to be charged too early (for example, a lower routine target due tomorrow plus a higher trip target due later gets planned as the higher level by tomorrow), which is a behavioral regression from the previous per-intent scheduling logic and increases unnecessary early charging.

Useful? React with 👍 / 👎.

chatgpt-codex-connector · 2026-02-18T09:41:00Z

providers/tesla/tesla-agent.ts

+              logVehicle(LogLevel.Warning, v, `Telemetry config blocked by scopes; backoff ${Math.round(delayMs / 60e3)}m`);
+            }
+          }
+          return;


Await queued work before returning on telemetry scope errors

The missing scopes handler returns immediately from serviceWork, bypassing the final await Promise.all(waitFor) that normally joins already queued async work. When this branch is hit, previously started vehicleWork/telemetry API calls run detached from the service loop, so their failures are no longer observed here and the next poll can start while prior work is still in flight.

Useful? React with 👍 / 👎.

Implemented charge windows instead of just picking random low price s…

f4aff26

…egments Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 requested a review from Copilot February 6, 2026 01:35

Copilot started reviewing on behalf of fredli74 February 6, 2026 01:35 View session

chatgpt-codex-connector bot reviewed Feb 6, 2026

View reviewed changes

server/logic.ts Show resolved Hide resolved

server/logic.ts Outdated Show resolved Hide resolved

Copilot AI reviewed Feb 6, 2026

View reviewed changes

fredli74 and others added 6 commits February 6, 2026 02:46

logic.ts trailing whitespace

f0cabe5

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update server/logic.ts

6ee0dd7

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Reuse now instead of Date.now()

d77163b

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Incorrectly removed current timeslot from price_data SQL

62ca0a6

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Inline score calculation for remainders.

2f49ce8

Track requested charge levels in status messages

56f2dbf

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 requested a review from Copilot February 6, 2026 02:08

Copilot started reviewing on behalf of fredli74 February 6, 2026 02:08 View session

Copilot AI reviewed Feb 6, 2026

View reviewed changes

server/logic.ts Outdated Show resolved Hide resolved

fredli74 added 2 commits February 6, 2026 20:00

Refactor soft intents: type definition and status function

dfdbcbc

Disable ESLint for v-html in alerts

289a128

fredli74 requested a review from Copilot February 6, 2026 19:10

Copilot started reviewing on behalf of fredli74 February 6, 2026 19:10 View session

chatgpt-codex-connector bot reviewed Feb 6, 2026

View reviewed changes

server/logic.ts Outdated Show resolved Hide resolved

server/logic.ts Outdated Show resolved Hide resolved

Copilot AI reviewed Feb 6, 2026

View reviewed changes

server/logic.ts Show resolved Hide resolved

app/src/app.vue Show resolved Hide resolved

app/src/app.vue Show resolved Hide resolved

app/src/app.vue Show resolved Hide resolved

server/logic.ts Show resolved Hide resolved

fredli74 added 4 commits February 7, 2026 02:35

Still do SoftIntents during manual charge, because of trip scheduling

edee93d

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

Fix backoff logic to always run at least once if we have a targetMs <…

cd6e37c

… quantumMs Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

Fix: Apply fill plan only if windows exist and time sufficient

a05c30e

Tesla-agent would set charge limit (wantedSoc) too low and risk the v…

9970b35

…ehicle not waking up to charge, cause it though the level was ok already Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 requested a review from Copilot February 7, 2026 01:51

Copilot started reviewing on behalf of fredli74 February 7, 2026 01:52 View session

Copilot AI reviewed Feb 7, 2026

View reviewed changes

app/src/app.vue Show resolved Hide resolved

app/src/app.vue Show resolved Hide resolved

Improve logging

abaa689

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

fredli74 added 6 commits February 12, 2026 01:58

Refactor charge scheduling to use dynamic programming instead of spli…

0ba6c3c

…tting logic.

Refactor vehicle logging to use dedicated logger

d6249df

Increase warmup penalty to 15 minutes and add scheduling window logging.

eff4b93

Fix existingPrecon check

e792fc0

Signed-off-by: Fredrik Lidström <fredli74@gmail.com>

Add error logging for vehicleWork failures

2469fd9

Add exponential backoff for telemetry config on 403 errors

8ec1159

fredli74 requested a review from Copilot February 18, 2026 09:32

Copilot started reviewing on behalf of fredli74 February 18, 2026 09:33 View session

Copilot AI reviewed Feb 18, 2026

View reviewed changes

chatgpt-codex-connector bot reviewed Feb 18, 2026

View reviewed changes

-        // DP over (startSlot, neededSteps). Gap penalty is modeled as extra score cost, not extra time.
-        const dpTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));
-        const dpWindows = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(Infinity));
-        const choiceTable = Array.from({ length: numSlots + 1 }, () => new Array(timeNeededSteps + 1).fill(null));
+        // Estimate DP memory usage and warn if very large.
+        const dpRows = numSlots + 1;
+        const dpCols = timeNeededSteps + 1;
+        const dpCellsPerTable = dpRows * dpCols;
+        const dpTablesCount = 3;
+        const totalDpCells = dpCellsPerTable * dpTablesCount;
+        const LARGE_DP_THRESHOLD = 500_000;
+        if (totalDpCells > LARGE_DP_THRESHOLD) {
+          vehicleLog(
+            LogLevel.Warn,
+            vehicle.vehicle_uuid,
+            `scheduleWindows(${scheduleTag}): large DP allocation: ` +
+            `slots=${numSlots} steps=${timeNeededSteps} cellsPerTable=${dpCellsPerTable} totalCells=${totalDpCells}`
+          );
+        }
+        // DP over (startSlot, neededSteps). Gap penalty is modeled as extra score cost, not extra time.
+        const dpTable = Array.from({ length: dpRows }, () => new Array(dpCols).fill(Infinity));
+        const dpWindows = Array.from({ length: dpRows }, () => new Array(dpCols).fill(Infinity));
+        const choiceTable = Array.from({ length: dpRows }, () => new Array(dpCols).fill(null));

	if (p.level !== undefined && (wantedSoc === undefined \|\| wantedSoc < p.level)) {
	if (wantedSoc === undefined && p.level !== undefined) {

		vehicleLog(LogLevel.Debug, input.id, `updateVehicle: ${JSON.stringify(input)}`);
		await context.db.getVehicle(accountFilter(context.accountUUID), input.id);

		if (i.beforeTs !== undefined) intentDeadline = Math.min(intentDeadline, i.beforeTs);
		if (i.level > intentMaxLevel) intentMaxLevel = i.level;

Conversation

fredli74 commented Feb 6, 2026

Uh oh!

fredli74 commented Feb 6, 2026

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

fredli74 commented Feb 6, 2026

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

fredli74 commented Feb 18, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!