Return local ai-dememory readiness checks without mutating files.

Return structured Markdown memory validation and conflict scan status without mutating files.

Search ranked local memory results from the SQLite index.

Read a memory document by id or repository-relative path.

Write a reviewed proposal candidate to inbox/llm-captures/.

Return token-budgeted memory context for an explicit query or generated working context.

Log that a memory search result was used.

Run secret scan, validation, and rebuild the SQLite index.

Generate a consolidation dry-run report. Does not mutate memories.

Scan selected paths or the repository for suspected secrets.

Return a lightweight graph of memory, tag, project, type, scope, and reference relationships.

Record good/bad usefulness feedback for a memory or last seen memory.

Return generated lifecycle scores from retrieval and outcome feedback.

Return safe sleep consolidation candidates without mutating memory.

Write selected sleep consolidation candidates to inbox/sleep-consolidation/.

Read the generated working/current.json snapshot if present.

Summarize current working state, recent-session file, and recent handoffs.

Write generated working/current.json and working/recent-session.md after secret scanning.

Write a generated working/handoffs/ Markdown handoff after secret scanning.

Write a reviewed recall miss candidate to inbox/recall-feedback/.

Report recall fixture provenance and reviewed-promotion freshness.

Check whether a query and expected memory are a recall miss candidate without writing files.

Report pending recall miss review work without writing fixture files.

Render the weekly recall review packet without writing reports, fixture files, or miss outcomes.

List generated recall review packet archives without writing files, fixtures, or miss outcomes.

Plan generated recall review packet archive pruning without deleting files, writing fixtures, or closing misses.

Reject or dismiss a reviewed recall miss without writing recall fixtures or canonical memory.

Report whether recall fixtures justify a future vector search experiment.

Audit durable memories for reviewed provenance metadata.

Show configured providers, schedule settings, generated artifacts, generated packet archives, recent reports, and lock state.

Import configured provider chat/session files into inbox/imports/ for review.

Capture explicit text or a repository-local file into inbox/imports/ for review.

Inspect local git history and preview review-first lesson candidates; dry_run=false writes into inbox/git-lessons/.

Run an opt-in daily or weekly maintenance profile.

Return platform scheduler commands and reviewed cron export entries without installing them.

Return configured schedule settings and platform status commands without querying or mutating the OS scheduler.

Check local scheduler, Docker, and crontab command availability without running commands.

Return reviewed manual release acceptance evidence status.

Return manual acceptance completion state without recording evidence.

Return remaining and blocked manual acceptance next actions without recording evidence.

Return a reviewed-evidence template and record command without recording acceptance evidence.

Render the manual acceptance review packet without writing reports or recording evidence.

List generated manual acceptance packet archives without writing files or recording evidence.

Plan generated manual acceptance packet archive pruning without deleting files or recording evidence.

Return local v2 release evidence without writing report files.

Render the v2 release evidence Markdown report without writing report files.

Plan manual TestPyPI or PyPI publishing without uploading packages.

Return read-only implementation status for the v2 operational roadmap phases.

Return supported local provider hook events.

Return local hook config fragments for Codex or Claude Code.

Return managed hook instruction install status without writing files.

Record a reviewed outcome on a hook capture under inbox/session-events/ without promoting memory.

Detect known local LLM provider folders for optional chat import.

Return configured provider import readiness without reading or importing chat files.

Return reviewable provider configure/import commands without mutating config or reading provider files.

Return review-first vault, MCP, provider, hook, and scheduler setup commands without mutating files.

Return combined validation, context config, manual acceptance, recall, vector, provider, scheduler, maintenance, and review health without running commands or writing files.

Return secret-scan findings with deterministic review ids and suppression metadata.

Return ignored false-positive suppressions whose current scanner finding no longer exists.

Record a reviewed secret-scan false-positive suppression in .ai-dememory-ignore.toml.

Record a reviewed false-positive unsuppression in .ai-dememory-ignore.toml.

Return duplicate, preference, project decision, and restricted memory conflict candidates.

Record that a conflict candidate was reviewed and intentionally dismissed.

Write a conflict merge proposal under inbox/conflict-resolution/ and audit the decision.

Record a reviewed keep decision for a conflict without editing canonical memory.

Return built-in review modes and the active configured mode.

Persist the active review mode in .ai-dememory.toml without editing canonical memory.

Return mode-specific LLM, human, and validation steps for a review kind.

Store an advisory LLM/client review recommendation under inbox/review-recommendations/ without applying it.

List advisory review recommendation artifacts without applying review outcomes.

List archived accepted/rejected advisory recommendation artifacts without moving files.

Preview restoring one archived advisory recommendation artifact to the active inbox without moving files.

Render the reviewed recommendation outcome sign-off report without writing files.

Record accepted/rejected status on an advisory review recommendation artifact without applying it.