Lovstudio Auto Context

AutoContext: Context Operator

Not just a health check — a full operator over everything that shapes context. Three layers by action sensitivity:

The harness owns /fork and /compact; this skill cannot invoke them. But it can do everything else and will.

Auto Mode (via Plugin Hook)

When the lovstudio plugin is enabled, a UserPromptSubmit hook monitors transcript size. Above threshold (40+ entries or 150KB+), it injects a lightweight <auto-context> reminder.

When you see `<auto-context>`:

Rules:

• 1 sentence max unless acting.
• If context is fine, say nothing.
• Never auto-fork/auto-compact (can't anyway — harness-only).
• Don't mention "AutoContext" unless asked.

Manual Mode (/lovstudio-auto-context [args])

Two call shapes:

A. Bare call — health report + opportunistic memory write

/lovstudio-auto-context

1. Measure — estimate turns, tool calls, distinct topics 2. Assess — healthy / getting noisy / polluted / critical 3. Scan recent turns for unpersisted feedback/preferences — if the user stated a rule or preference earlier in the session that should persist to future conversations (typical trigger phrases: "从今以后", "以后都", "所有 X 应该 Y", "别再", "记住"), and no memory was written, auto-execute: write the memory file + update MEMORY.md. Report the path. 4. Recommend harness actions if needed (/fork, /compact, /btw) with the exact command to paste.

Keep to 3-5 lines unless taking confirm-first actions.

B. With arguments — targeted context operation

/lovstudio-auto-context <free-form instruction>

Parse the instruction and route to the right action class:

Action: Write Project Memory (auto-execute)

Directory: <resolved-agent-home>/projects/<project-slug>/memory/. Resolve <resolved-agent-home> from the User Configuration section before writing. The memory directory should already exist; do not create a new runtime layout silently.

Procedure: 1. Pick filename: <type>_<topic>.md (e.g. feedback_output_paths.md). 2. Write with the required frontmatter (name, description, type). 3. For feedback / project types, include **Why:** and **How to apply:** lines. 4. Update MEMORY.md with one-line pointer under 150 chars. 5. Report the absolute + relative path.

Action: Edit Global CLAUDE.md (confirm-first)

Target: <resolved-agent-home>/CLAUDE.md.

Procedure: 1. Read the file. 2. Locate the best section for the addition (match existing heading like "输出规范", "网络 / 代理", "Debugging Discipline"; or create a new section if none fits). 3. Show the proposed diff as a fenced block with -/+ lines. 4. Ask: "执行这个修改？(yes/no)" 5. On "yes" → apply via Edit. On anything else → abort, optionally offer to save as project memory instead.

Never skip the diff step. CLAUDE.md is user-authored authoritative config; silent edits erode trust.

Action: Edit Project CLAUDE.md (confirm-first)

Same flow as global, but target is <cwd>/CLAUDE.md (walk up if not at root). Same diff-then-confirm requirement.

Action: Suggest Harness Commands (suggest-only)

Produce the exact string the user should paste, not a description:

context polluted — paste this to fork:
/fork

or

approaching capacity — compact first, then continue:
/compact

Do not wrap in explanations. The command is the deliverable.

What this skill cannot do (be honest about it)

• Invoke /fork, /compact, /btw, /clear, new session — harness-only.
• Edit another agent's transcript.
• Auto-install hooks — use /update-config for that.

If the user asks for one of these, say so and give them the command to paste.

Output convention

Every action that writes a file must echo its absolute + relative path (per project-wide output-paths rule). Applies to memory files, CLAUDE.md edits (show final path), and any derived artifacts.

User Configuration

Resolve agent paths in this order:

1. LOVSTUDIO_AUTO_CONTEXT_AGENT_HOME. 2. Shared profile ${LOVSTUDIO_SKILLS_PROFILE:-$HOME/.lovstudio/skills/profile.json} keys: agent.home, claude.home, or runtime.agent_home. 3. Ask the user once for the agent home directory and use that value for the current operation.

For alternate runtimes, set LOVSTUDIO_AUTO_CONTEXT_AGENT_HOME explicitly or add the relevant profile key.

Changelog

[0.3.0] - 2026-05-07

Added

• resolve agent home through env or profile
• remove fixed runtime directory default from AutoContext docs
• document profile keys for compatible agent runtimes

[0.2.1] - 2026-05-07

Fixed

• standardize agent home configuration
• add public install command and user configuration docs

0.2.0 — 2026-04-18

• Expand scope from "hygiene checker" to context operator covering all

context-affecting actions the skill can reach.

• Add action layer 1 (auto-execute): write project memory when unpersisted

feedback/preferences are detected in the session.

• Add action layer 2 (confirm-first): edit ~/.claude/CLAUDE.md or project

CLAUDE.md with diff-preview-then-confirm flow.

• Add action layer 3 (suggest-only): surface /fork, /compact, /btw as

exact-paste strings. Document that harness owns these — skill can't invoke them.

• Add argument-parsing for targeted ops: "记到全局", "记到项目", "记住 X",

"忘掉 X", "该分叉了吗".

• Output-path convention: every file-writing action echoes absolute +

relative path.

0.1.0

• Initial release: hook-triggered <auto-context> nudge + manual

/lovstudio-auto-context health report.

AutoContext

Context operator for Claude Code sessions — not just a health check.

What It Does

Three capability layers, routed by action sensitivity:

• Auto-execute: writes project memory when it spots unpersisted

feedback/preferences; updates MEMORY.md index.

• Confirm-first: edits the configured global instructions file or project CLAUDE.md

with a shown diff and explicit "yes" before applying.

• Suggest-only: harness commands like /fork, /compact, /btw —

surfaced as exact-paste strings.

You (turn 31):  "从今以后所有输出都要带路径"
                 ↑ AutoContext writes feedback memory automatically

You (turn 45):  "/lovstudio-auto-context 记到全局"
                 ↑ Shows diff of proposed global instructions edit, waits for yes

You (turn 80):  transcript size crosses threshold
                 ↑ Suggests: paste `/fork` (harness owns this one)

Install

npx lovstudio skills add auto-context -g -y

Works standalone as a manual skill. Auto-trigger on long transcripts requires the lovstudio plugin which registers the UserPromptSubmit hook.

Manual Use

/lovstudio-auto-context                              # health report + opportunistic memory write
/lovstudio-auto-context 记到全局                      # edit configured global instructions with confirm
/lovstudio-auto-context 记到项目                      # edit project CLAUDE.md with confirm
/lovstudio-auto-context 记住 X                        # write project memory
/lovstudio-auto-context 该分叉了吗                    # evaluate + suggest /fork

What It Cannot Do

Harness-owned commands (/fork, /compact, /btw, /clear, new session) can't be invoked programmatically. The skill will print the exact command for you to paste.

User Configuration

Set LOVSTUDIO_AUTO_CONTEXT_AGENT_HOME or configure ${LOVSTUDIO_SKILLS_PROFILE:-$HOME/.lovstudio/skills/profile.json} with agent.home, claude.home, or runtime.agent_home. If none is available, the skill asks once before editing global files.

Version

0.3.0 — resolves agent home through env/profile/user input instead of a fixed runtime directory. See CHANGELOG.md.

License

MIT