Give your coding agents a manager.

Skip keeps all your Claude Code, Codex and OpenCode sessions moving, coordinated and on track. You stay focused on what matters.

API pagination

Claude Code

✓12 tests passed✓Reviewing edge cases...⠋Thinking about cursor boundaries...

Usage metering

Codex

✓Inspecting usage event schema...✓rg "usage_events" Sources⠸Read Metering/UsageStore.swift

Passwordless login

Claude Code

✓Search existing rate limit policy✓Edit Passwordless.swift (+24 -3)⠦Run authentication tests

Session migration

Codex

✓Run migration fixtures✓Comparing 184 migrated records...⠏Preparing agent review

Skip is coordinatingFour agents working.I’ll let you know when they need you.

→

Backed by

Skip keeps sessions moving

Work with Skip to set the plan and keep every session moving toward the goal.

ProjectStatusNext up

API paginationWorkingAgent question

Usage meteringWorkingBackfill

Passwordless loginWorkingRate limit

Session migrationWorkingReview

Dependency auditWorkingReview

Search indexingWorkingIndex documents

CI failuresWorkingRetry checks

Webhook retriesWorkingChoose policy

You step in when judgement is needed

Skip batches the questions your agents cannot answer, brings the context, and lets you steer the work in one move.

SkipWebhook retries

↗Opened Webhook retries

The retry policy needs one product decision.

Did we cover this in the brief?

product-brief.mdRetry webhooks so deliveries do not get missed.Skip recommends a 24 hour cutoff.

Message Skip… ↑

Webhook retry policyCodex · live session

needs direction

✓Read Webhooks/RetryPolicy.swift

⇄Compared existing delivery guarantees

The brief does not specify whether failed webhooks should retry for 24 or 72 hours. Which policy should I implement?

✓ Direction sent. The session is moving again.

Keep using your favorite tools.

Move between Skip, your terminal, and the Codex and Claude Mac apps without losing the session, the decisions, or the plan.

▱Webhook retry policy•••Open⌄Commit⌄

Codex

✓ Restored session context

Implementer

The brief doesn’t specify whether failed webhooks should retry for 24 or 72 hours. Which policy should I implement?

We should use 24 hours as the retry window.

Got it. I’ll update the endpoint and affected tests.

Add a regression test for the retry cutoff too.

Updating retry policy and regression test…

Message Implementer… ↑

Skip reminds you what you were trying to do.

Once Skip understands the goal, it keeps what you want and what your agents are building in sync.

Project briefPasswordless login

INTENTS

Remove the password field

Reduce reset tickets

Keep security review clean

CONSTRAINTS

Nobody gets logged out

No new paid dependency

Links expire in 10 minutes

Skip is checking the workBrief and implementation stay in sync.

live

✓Passwordless onlyAligned

·No paid dependencyChecking

·Links expire in 10 minutesChecking

·No forced logoutChecking

All clearThe implementation still matches the plan.

Your whole team works from the same picture.

See what everyone’s agents are doing, know who owns each piece of work, and stop two sessions from solving the same problem. Every useful prompt and decision becomes something the team can build on.

Who is running what4 teammates · 7 live agent sessions

Live

MinaPlatform

2 sessions

API paginationTesting cursor boundaries

Running

CI failuresTracing the flaky suite

Investigating

AveryProduct

2 sessions

Webhook retriesImplementing the 24 hour cutoff

Writing

Release notesSummarizing merged changes

Drafting

JulesInfrastructure

2 sessions

Usage meteringBackfilling event totals

Reviewing

Search indexingChecking memory profile

Running

SamIdentity

1 session

Passwordless loginVerifying session expiry

Testing

Shared learning

A useful decision is ready for the next teammate.

Prompts, decisions, and tests stay attached to the work so every agent starts smarter.