Follow Discord
Sweep 10 Nov 2026 · 02:18Z Build v2.1.322 539 read Stable v2.1.311 Latest v2.1.322 Next v2.1.322 Feeds RSS JSON llms.txt llms-full.txt Unofficial
Prompt changes · v2.1.322 against v2.1.321

1 change since v2.1.321, landing in 4 of 20 prompts.

Every prompt v2.1.322 sent, compared word for word with the capture before it. Each card is one edit, with every prompt it reached; the table after them shows where each one landed.

What changed 1

1

System prompt · under “Harness”

New rule keeps user's personal details out of requests to outside services

Claude is now told to treat what it learns about the user and their project as private and never put their name, project, email or other personal details in headers, URLs or payloads sent to external services unless the user explicitly asks. If a service wants contact details from automated clients, such as an email in the User-Agent, Claude should use a generic client name or ask the user. The addition appeared in 3 of 5 captures.

Behind a feature flag

+76 −0 words

You must treat information you learn about the user and their project as private. Never include the user's name, project, email address or other personal details in requests to outside services, such as in a header, URL, or payload, unless the user explicitly asks you to do so. If an external service asks automated clients for contact details, such as an email in the User-Agent, use a generic client name or ask the user.

Where each change landed

Every prompt the release sent, one row each. A number is a change from the list above; click it to jump there. Rows with more than one model string count them, because each model gets its own copy of the prompt.

Prompt 1 Status
Main prompt
Claude Code, default27 models · unchanged
Claude Code, seeded with this account's feature flags27 models 27/27 1 change
Claude Code, seeded with anonymous feature flags27 models 27/27 1 change
Agent SDK, default27 models · unchanged
Agent SDK, seeded with this account's feature flags27 models 27/27 1 change
Agent SDK, seeded with anonymous feature flags27 models 27/27 1 change
Slash commands
/init1 model · unchanged
/security-review1 model · unchanged
/team-onboarding1 model · unchanged
Compaction
Compaction instruction1 model · unchanged
System reminders
Claude md1 model · unchanged
Hook pretool block1 model · unchanged
Read empty1 model · unchanged
Read offset1 model · unchanged
Skill run1 model · unchanged
Todo nag1 model · unchanged
Toolsearch deferred1 model · unchanged
Subagents
explore1 model · unchanged
general-purpose1 model · unchanged
plan1 model · unchanged

Did not change

31 tool descriptions are word for word the same.

Agent AskUserQuestion Bash CronCreate CronDelete CronList DeferredToolPlaceholder Edit EnterPlanMode EnterWorktree ExitPlanMode ExitWorktree ListAgents Monitor NotebookEdit PushNotification Read ReportFindings ScheduleWakeup SendMessage Skill TaskCreate TaskGet TaskList TaskStop TaskUpdate ToolSearch WebFetch WebSearch Workflow Write

16 prompts did not move at all: Claude Code, default, Agent SDK, default, Slash commands: /init, Slash commands: /security-review, Slash commands: /team-onboarding, Compaction: Compaction instruction, System reminders: Claude md, System reminders: Hook pretool block, System reminders: Read empty, System reminders: Read offset, System reminders: Skill run, System reminders: Todo nag, System reminders: Toolsearch deferred, Subagents: explore, Subagents: general-purpose, Subagents: plan.

Set aside as capture noise

These change on every capture by construction, so they are masked before comparing and never counted as a change.

  • 352 billing header lines
  • 220 capture date lines
  • 220 host kernel lines
Feedback