agentel 0.2.0 → 0.2.3

package/README.md

@@ -18,8 +18,8 @@ Core capabilities:
   coding agents
 - local web viewer for full conversation history inspection
 - device-scoped S3-compatible upload sync for backup and future multi-device recall
-- lifecycle, config, autostart, and MCP config helper commands
-- local reset command for starting init from a clean archive/config state
+- lifecycle, config, watcher login, and MCP config helper commands
+- local update/reset commands for rebuilding archives after package changes
 
 The local HTTP OTLP collector, S3-compatible remote sync, and team deployment
 surfaces are configurable extension points for environments that install those
@@ -31,7 +31,7 @@ Install the CLI globally so agent-facing recall commands can call `agentlog`
 from PATH:
 
 ```sh
-npm install -g agentlogs
+npm install -g agentel
 agentlog init
 ```
 
@@ -41,21 +41,21 @@ ref for repeatable installs:
 ```sh
 npm install -g brianlzhou/agentlog
 # or
-npm install -g brianlzhou/agentlog#v0.2.0
+npm install -g brianlzhou/agentlog#v0.2.3
 agentlog init
 ```
 
 You can also install it into a project and run it through npm/npx:
 
 ```sh
-npm install agentlogs
-npx agentlogs init
+npm install agentel
+npx agentel init
 ```
 
 Or run the published package without keeping a local install:
 
 ```sh
-npx agentlogs init
+npx agentel init
 ```
 
 Requirements:
@@ -124,21 +124,21 @@ npm run smoke:pack
 ## Recall
 
 ```sh
-agentlog recall add-to codex
-agentlog recall add-to claude
-agentlog recall add-to gemini
-agentlog recall add-to antigravity
-agentlog recall add-to devin
-agentlog recall add-to cursor
-agentlog recall add-to cline
-agentlog recall add-to opencode
-agentlog recall add-to aider
+agentlog integrations add-to codex
+agentlog integrations add-to claude
+agentlog integrations add-to gemini
+agentlog integrations add-to antigravity
+agentlog integrations add-to devin
+agentlog integrations add-to cursor
+agentlog integrations add-to cline
+agentlog integrations add-to opencode
+agentlog integrations add-to aider
 ```
 
-`recall add-to codex` installs the MCP server plus a Codex recall skill.
-`recall add-to claude` installs the MCP server plus a Claude `/recall`
+`integrations add-to codex` installs the MCP server plus a Codex recall skill.
+`integrations add-to claude` installs the MCP server plus a Claude `/recall`
 command plus a Claude skill under `~/.claude/skills/agentlog-recall/SKILL.md`.
-`recall add-to gemini` installs the MCP server plus a Gemini `/recall` command
+`integrations add-to gemini` installs the MCP server plus a Gemini `/recall` command
 under `~/.gemini/commands/recall.toml`.
 
 Additional native surfaces are installed where the client exposes one:
@@ -183,7 +183,7 @@ agentlog-recall
 ## Daily CLI
 
 `agentlog status`, `agentlog history`, `agentlog show`, `agentlog import`, `agentlog index`, and
-`agentlog autostart` use styled human-readable output by default. Use `--json`
+`agentlog watcher login` use styled human-readable output by default. Use `--json`
 on `status`, `history`, and import status/results for script-friendly
 structured output. `agentlog status` includes the most recently archived
 conversation threads and the sources currently monitored by the supervisor;
@@ -197,17 +197,23 @@ agentlog history "codebase explanation" --provider codex-cli --limit 10
 agentlog history --repo github.com/acme/widgets --since 90d
 agentlog show <session-id>
 agentlog show <session-id> --path
-agentlog history --web
-agentlog history --web --no-open
+agentlog web
+agentlog web --no-open
 ```
 
-`agentlog history --web` starts the local conversation viewer and opens it in
+`agentlog web` starts the local conversation viewer and opens it in
 your default browser. Use `--no-open` when you only want to print the local URL
 and keep the server running. The left rail is a
 repo tree sorted by the latest updated session, each folder pages through
 sessions with a load-more control, search results reuse the same session
 reader, and the transcript pane can switch between readable chat bubbles and
-the raw markdown archive. The static viewer uses shadcn/ui-style design tokens
+the raw markdown archive. Readable chat bubbles load from pre-baked
+`*.view.json` payloads; raw markdown is fetched only when source view is
+opened. Search runs as you type with a short debounce. The web endpoint uses a
+warm compatible index when available, but it will not synchronously parse or
+rebuild an obsolete index during an interactive query or scan every rendered
+conversation as a fallback; rebuilds are left to `agentlog index rebuild` or
+the supervisor. The static viewer uses shadcn/ui-style design tokens
 and compact button/input/select/sidebar patterns without requiring a frontend
 build step. Archives still keep stable `path:<hash>` keys for folders without
 git identity, but the UI displays the local path.
@@ -218,11 +224,17 @@ Provider filters use one stable order: OpenAI (`codex-cli`, `codex-desktop`,
 (`devin-cli`), then other local tools (`cursor`, `cline`, `opencode`,
 `aider`).
 
-When running, the supervisor polls the watcher source list chosen during init
-every 30 seconds using the configured import window, defaulting to the last 30
-days. The default watcher choices are Codex CLI, Codex Desktop, Claude Code CLI,
-Claude Code Desktop, Claude Workspace, Gemini CLI, Antigravity, Devin CLI,
-Cursor, Cline, OpenCode, and Aider. New configs still support
+The supervisor is agentlog's local background watcher. When it is running, it
+polls the watcher source list chosen during init every 30 seconds using the
+configured import window, defaulting to the last 30 days. If you opt out of
+starting it at login, agentlog does not install a login item. During setup,
+uncheck `Start watcher at login` or run `agentlog init --no-autostart`; you can
+still run `agentlog import --source all` for a one-time catch-up,
+`agentlog watcher start` to watch for the current session, or
+`agentlog watcher login enable` later. The default
+watcher choices are Codex CLI, Codex Desktop, Claude Code CLI, Claude Code
+Desktop, Claude Workspace, Gemini CLI, Antigravity, Devin CLI, Cursor, Cline,
+OpenCode, and Aider. New configs still support
 `imports.autoDiscoverSources=true`, but init records the chosen watcher list
 exactly by setting `imports.autoDiscoverSources=false`.
 Cursor raw SQLite recovery is intentionally left to explicit imports such as
@@ -233,20 +245,21 @@ multi-session stores such as Cursor SQLite and Devin `sessions.db`, lives in
 [`docs/history-source-handling.md`](docs/history-source-handling.md).
 
 Remote sync can be configured during init or from the CLI. Local storage remains
-canonical. Current sync is upload-only: this machine pushes changed archive
+canonical. Normal sync is upload-only: this machine pushes changed archive
 objects to the remote target and never deletes objects from the bucket just
 because they are missing locally.
 
 ```sh
-agentlog sync \
-  --endpoint https://<account-id>.r2.cloudflarestorage.com/<bucket> \
-  --device-name "work-laptop" \
-  --access-key-id <id> \
-  --secret-access-key <key>
+agentlog sync configure
+agentlog sync
 ```
 
-If you prefer to split the values manually, use
-`--endpoint https://<account-id>.r2.cloudflarestorage.com --bucket <bucket>`.
+`agentlog sync configure` opens a picker in a terminal. Choose an existing
+configured/env remote, or choose the last option to configure a new R2, S3,
+S3-compatible, or local-folder target. After a target is selected, the CLI asks
+what to do next: sync now, create a snapshot, change autosync cadence, or exit.
+Non-interactive scripts can still pass flags such as `--endpoint`, `--bucket`,
+and `--access-key-id`.
 
 The same values can be supplied with `AGENTLOG_REMOTE_ENDPOINT`,
 `AGENTLOG_REMOTE_BUCKET`, `AGENTLOG_REMOTE_ACCESS_KEY_ID`, and
@@ -264,9 +277,9 @@ the bucket or `agentlog/` prefix with `ListBucket` and object read/write
 permissions.
 
 During `agentlog init`, a configured cloud destination is uploaded immediately
-after the initial backfill. If auto-start is enabled, init also asks for a cloud
-autosync cadence; the background supervisor then uploads new archive changes on
-that interval.
+after the initial backfill. If the watcher is allowed to start at login, init
+also asks for a cloud autosync cadence; the supervisor then uploads new archive
+changes on that interval whenever it is running.
 
 Remote objects are namespaced by device so multiple machines can upload into
 one bucket without overwriting each other's archive roots:
@@ -284,13 +297,40 @@ s3://<bucket>/agentlog/
         sessions/...
 ```
 
-Use `agentlog snapshot` to upload a redundant point-in-time copy under
-`agentlog/snapshots/<timestamp>/<device>/...`:
+In a terminal, `agentlog sync` asks you to choose the remote target, previews
+the upload-only plan, and requires a confirmation phrase before writing. This is
+intentional even when only one remote is configured, because multiple remotes
+can exist in config/env and a normal sync can overwrite same-key objects for the
+selected device namespace. Use `--dry-run` to preview in scripts, or `--yes` to
+skip the guided confirmation.
+
+Use `agentlog sync snapshot` to upload a redundant point-in-time copy under
+`agentlog/snapshots/<timestamp>/<device>/...`. In a terminal it asks you to
+choose the remote, lists existing snapshots, asks for the snapshot name, previews
+the write, and then confirms before uploading:
+
+```sh
+agentlog sync snapshot
+```
+
+If the remote copy has stale objects after a local reset, cleanup, or full
+reimport, use the explicit replace path. In a terminal it previews the selected
+remote, shows how many remote objects will be deleted, and requires typed
+confirmation before it deletes this device namespace under
+`agentlog/devices/<device>/` and uploads the current local archive.
 
 ```sh
-agentlog snapshot --name before-reimport
+agentlog sync replace
+agentlog sync wipe
 ```
 
+`agentlog sync wipe` is delete-only. It asks which remote to use, asks for the
+scope (`device`, one `snapshot`, all `snapshots`, `prefix`, or `bucket`),
+previews the exact target and prefix, then requires a two-step typed
+confirmation. `device` leaves snapshots untouched; `snapshots` leaves normal
+device uploads untouched. Run `agentlog sync` afterward to upload this device
+again. Scripts can use `--dry-run` or `--yes`.
+
 Receive-only and two-way sync are intentionally not active yet. The intended v1
 shape is to read other device namespaces and merge normalized session metadata
 without treating absence on one machine as a deletion. Remote deletes should be
@@ -328,14 +368,25 @@ the same database into every session folder.
 `tool.completed`. Parser versions are stamped by source type so importer output
 changes can trigger reimport with a new fingerprint. Recall/search builds a
 keyword index over event text first and falls back to transcript/markdown for
-legacy archives without events.
+legacy archives without events. The local search index stores compact term
+postings for CLI compatibility plus a SQLite FTS5 sidecar for fast web
+queries; when either index format changes, `agentlog history` and `agentlog
+index` rebuild it from archived transcripts/events without a full source
+reimport. The web viewer avoids doing that rebuild on a keystroke so a large
+old index, or a full-archive Markdown fallback, cannot block interactive
+search.
 
 Stats are import-time metadata, not viewer-time transcript repair. Archive
 metadata stores message counts, user-message counts, token usage, and models for
-each session, and the web stats view reads those fields directly. During pre-v1
-development, if those stats fields or parser semantics change, rebuild the local
-archive with `agentlog reset --yes`, `agentlog init`, and
-`agentlog import --source all --since all`.
+each session, and the web stats view reads those fields directly. Cursor sessions
+without provider-reported token usage can also carry separately labeled
+`estimatedUsage`, which the stats view includes while reporting estimated token
+coverage. ChatGPT and Claude.ai exports without provider usage get estimated
+`metadata.usage` on their native chat messages, split into non-assistant input,
+assistant output, and Claude thinking output where the export provides separate
+parts. During pre-v1 development, if those stats fields or parser semantics
+change, rebuild the local archive with
+`agentlog update --yes --since all`.
 
 ChatGPT and Claude.ai web exports are imported manually from an official `.zip`,
 an unzipped export folder, or a direct JSON file. These imports are stored as
@@ -343,9 +394,11 @@ local scoped web-chat archives and displayed through virtual conversation roots
 such as `[chatgpt]conversations/<account-id>` and
 `[claude]conversations/<account-id>/<project>`. The importer records account
 metadata in `~/.agentlog/state/web-accounts.json`; use
-`agentlog accounts list` to inspect mappings and
-`agentlog accounts rename <provider> <account-id-or-username> --display-name <name>`
-to change the viewer display name. Repeated manual uploads are incremental:
+`agentlog import accounts list` to inspect mappings and
+`agentlog import accounts rename <provider> <account-id-or-username> --display-name <name>`
+to change the viewer display name. Claude.ai exports preserve conversation
+summaries and split structured thinking parts from visible assistant answers
+when the export includes that detail. Repeated manual uploads are incremental:
 unchanged conversations are skipped, and updated conversations replace the
 stable session for that provider/account/conversation id. Existing malformed
 pre-v1 web-chat archives are not migrated automatically; reimport from the
@@ -364,11 +417,50 @@ local home, config, state, cache, logs, and archive objects. Source application
 histories and MCP configuration files are not changed. Add `--keep-autostart`
 to leave the login item in place.
 
+Use `agentlog update` after installing a newer npm package when you want the new
+importer/parser logic to rebuild the local archive without redoing setup:
+
+```sh
+npm install -g agentel@latest
+agentlog update --yes --since all
+```
+
+`agentlog update` preserves `config.json`, redaction settings, web account
+labels, source histories, and recall integrations. It removes derived local
+archive, import, index, cache, and sync bookkeeping, then reimports configured
+local sources from the stored preferences. It does not touch remote sync objects
+by default; use `agentlog sync replace` when the remote should match the rebuilt
+local archive. It also does not rediscover manual ChatGPT/Claude.ai export
+files; reimport those web exports from the original ZIP/folder when needed.
+
+Use `agentlog config` to change `~/.agentlog/config.json` without rerunning the
+init wizard:
+
+```sh
+agentlog config path
+agentlog config setup
+agentlog config setup --watch-sources codex-cli,cursor --default-since-days 90
+agentlog config setup --sync-interval-minutes manual --no-autostart
+agentlog config set imports.defaultSinceDays 90
+agentlog config set sync.intervalMinutes manual
+agentlog config sources edit
+agentlog config sources set codex-cli,cursor,cline
+agentlog config sources add gemini-cli
+agentlog config sources remove claude-workspace
+```
+
+`agentlog config setup` reopens the preferences parts of init, including what
+the watcher polls, the rolling import window, autosync cadence, and login
+startup. It does not run an import.
+`agentlog config sources edit` opens just the watcher source picker from init,
+with the current config preselected.
+
 ## Import Windows
 
 `agentlog init` starts with interactive setup: choose archive destinations,
-choose the full local archive/cache path, choose background service behavior and
-recall commands or skills, then discover sources and optionally backfill history.
+choose the full local archive/cache path, choose whether the local watcher starts
+at login, and install recall commands or skills, then discover sources and
+optionally backfill history.
 After backfill, init asks which sources the background watcher should keep
 polling, then offers local OTel bridges for Claude Code, Gemini CLI, and Cline.
 Local archive storage is always enabled; R2, S3, and custom remote sync targets
@@ -376,10 +468,12 @@ can be added as upload-only optional destinations with a device name for the
 remote namespace. Discovery and import phases show progress bars while they scan
 local stores.
 
-After discovery, init offers a checkbox-style source picker. Claude SDK jobs
-are shown as a separate opt-in source because batch SDK traffic can exceed
-interactive sessions. The selected sources are saved in config and used by
-later `agentlog import --source all` runs unless `--sources` is provided
+After discovery, init offers a checkbox-style source picker. Rows marked `[x]`
+are selected; type one or more row numbers, such as `1 3 8`, to toggle sources
+on or off, then press Enter with no input to accept the current selection.
+Claude SDK jobs are shown as a separate opt-in source because batch SDK traffic
+can exceed interactive sessions. The selected sources are saved in config and
+used by later `agentlog import --source all` runs unless `--sources` is provided
 explicitly.
 
 Default init sources:
@@ -390,7 +484,7 @@ Default init sources:
 - Claude Code Desktop metadata and Claude Workspace/local-agent sessions from
   the Claude app data, shown as separate toggles
 - Gemini CLI saved chats/checkpoints under `~/.gemini/tmp`, plus session/export JSONL stores with tool, usage, and checkpoint metadata
-- Antigravity task/plan/walkthrough artifacts under `~/.gemini/antigravity/brain`
+- Antigravity task/plan/walkthrough artifacts under `~/.gemini/antigravity/brain`, plus partial trajectory summaries from Antigravity app state when no readable artifacts exist
 - Devin for Terminal sessions from `~/.local/share/devin/cli/sessions.db`
 - Cursor chats from older workspace `state.vscdb` SQLite stores and global
   `cursorDiskKV` Composer/Agent rows, including `aiService` prompt/generation
@@ -408,9 +502,13 @@ standard Claude Code transcript, so agentlog imports the transcript when it exis
 and only imports Claude Workspace sessions that contain actual prompt
 content.
 
-Windsurf import is disabled for now. Current Cascade transcripts are encrypted
-binary stores, so agentlog can detect that a session exists but cannot archive
-readable conversation text from it yet.
+Windsurf local cache scanning is disabled for now. Current Cascade transcripts
+are encrypted binary stores, so agentlog can detect that a session exists but
+cannot archive readable conversation text from that cache. Use Windsurf's
+"Download trajectory" Markdown export with `agentlog import windsurf <path>` to
+archive readable Cascade output. If you bulk-export multiple trajectories to a
+folder, import the folder directly, for example
+`agentlog import windsurf ~/windsurf-cascade-export`.
 
 For the full source-by-source implementation map, see
 [History Source Handling](docs/history-source-handling.md).
@@ -447,6 +545,6 @@ agentlog import --source cline --since all
 agentlog import --source opencode --since all
 agentlog import --source aider --since all
 agentlog import --source claude-sdk --since all
-agentlog accounts list
-agentlog accounts rename claude-web you --display-name "Personal Claude"
+agentlog import accounts list
+agentlog import accounts rename claude-web you --display-name "Personal Claude"
 ```

package/agentlog-spec.md

@@ -1,6 +1,6 @@
 # agentlog — spec v0.2
 
-A weekend-buildable archive and recall layer for agent coding sessions across Codex, ChatGPT exports, Claude, Gemini, Antigravity, Devin, and Cursor. Local-first, optionally cloud-backed via any S3-compatible storage. Web chats from ChatGPT and Claude.ai are importable via their official export flows. Windsurf is disabled until Cascade exposes a readable local transcript path.
+A weekend-buildable archive and recall layer for agent coding sessions across Codex, ChatGPT exports, Claude, Gemini, Antigravity, Devin, Cursor, and Windsurf trajectory exports. Local-first, optionally cloud-backed via any S3-compatible storage. Web chats from ChatGPT and Claude.ai are importable via their official export flows. Windsurf's encrypted local Cascade cache remains disabled, but downloaded trajectory Markdown is importable.
 
 ## What it does
 
@@ -78,18 +78,20 @@ Three layers, separable, in priority order:
 
 ## Auto-start at login
 
-`init` offers to install a platform-native login agent. Default: yes, with explicit opt-out.
+`init` offers to install one platform-native login item for the local watcher. Default: yes, with explicit opt-out.
 
 **Prompt during init:**
 
 ```
-agentlog can start automatically when you log in.
+Background Watcher
 
-This runs a small background service (~150MB RAM) that captures
-your agent conversations as they happen. You can stop it anytime
-with `agentlog stop` or remove auto-start with `agentlog uninstall`.
+The supervisor is agentlog's local watcher: it imports new history,
+refreshes indexes, and runs scheduled cloud sync.
+Leave this checked to install one login item. Uncheck it for manual-only mode.
 
-Start agentlog automatically at login? [Y/n]
+  1  [x] Start watcher at login
+      installs one user-level login item; unchecked means no continuous
+      watching until you run agentlog watcher start
 ```
 
 **Per-platform implementation:**
@@ -98,14 +100,14 @@ Start agentlog automatically at login? [Y/n]
 - **Linux:** systemd user unit at `~/.config/systemd/user/agentlog.service`. `Type=simple`, `Restart=on-failure`, `RestartSec=5`. Enabled with `systemctl --user enable --now`. Init detects whether `loginctl enable-linger` is needed and prompts separately: "Keep agentlog running when you're logged out? [y/N]" — default no.
 - **Windows:** Scheduled Task triggered at logon (`schtasks /create /tn "agentlog" /tr ... /sc onlogon`). Service-based install is a v1 enhancement.
 
-**Critical detail:** the launch agent runs `agentlog start --foreground`, not `agentlog start`. Foreground mode keeps the OS supervisor (launchd/systemd/Task Scheduler) as the parent — detaching breaks process tracking and crash restart.
+**Critical detail:** the launch agent runs `agentlog watcher start --foreground`, not `agentlog watcher start`. Foreground mode keeps the OS supervisor (launchd/systemd/Task Scheduler) as the parent — detaching breaks process tracking and crash restart.
 
 **Lifecycle commands:**
 
 ```
-agentlog autostart enable     # writes the launch agent/unit/task
-agentlog autostart disable    # removes auto-start, keeps agentlog installed
-agentlog autostart status     # shows current state
+agentlog watcher login enable     # writes the launch agent/unit/task
+agentlog watcher login disable    # removes auto-start, keeps agentlog installed
+agentlog watcher login status     # shows current state
 agentlog uninstall [--keep-data]
 ```
 
@@ -175,7 +177,7 @@ every source type. After release, when parser output changes for the same raw
 input, the source-type version must be bumped in the same change so stale
 archives can be replaced.
 
-**Migration:** `agentlog migrate --to <backend>` records the remote target; `agentlog sync` uploads the same markdown-primary object layout to any S3-compatible target under `agentlog/devices/<device-name>/...`. Local→R2 is a one-shot upload and then an incremental supervisor upload. Sync does not delete remote objects. Receive-only and two-way sync should read other device namespaces and merge normalized archive metadata without interpreting absence on one device as a delete. `agentlog snapshot` writes redundant point-in-time copies under `agentlog/snapshots/<timestamp>/<device-name>/...`.
+**Migration:** `agentlog sync configure` records the remote target through an interactive picker in terminals, while non-interactive scripts can still pass `--target`, `--endpoint`, and credentials. Choosing an existing remote opens useful next actions instead of only echoing the current config. `agentlog sync` uploads the same markdown-primary object layout to any S3-compatible target under `agentlog/devices/<device-name>/...`; terminal runs pick a remote, preview the upload-only plan, and confirm before writing. Local→R2 is a one-shot upload and then an incremental supervisor upload. Normal sync does not delete remote objects. `agentlog sync replace` is the explicit repair path: it previews the selected remote, requires typed confirmation, deletes only the current device namespace, and then uploads the current local archive. `agentlog sync wipe` is delete-only, asks for remote and scope, previews the chosen target and prefix, requires typed confirmation, and is followed by `agentlog sync` when the user wants to rebuild a remote copy from local state. Wipe scopes include the current device namespace, one snapshot, all snapshots, the configured prefix, and the bucket/root. Receive-only and two-way sync should read other device namespaces and merge normalized archive metadata without interpreting absence on one device as a delete. `agentlog sync snapshot` lists existing snapshots, asks for a name, confirms, and writes redundant point-in-time copies under `agentlog/snapshots/<timestamp>/<device-name>/...`.
 
 ## Repo keying
 
@@ -219,7 +221,7 @@ allowlist_repos:
 
 Each session gets a `redaction_summary` span: counts by category, no content. Users can audit "did this leak anything" without seeing what leaked.
 
-`agentlog reveal <session-id>` re-renders un-redacted from local cache. Local-only — never works on remote/team archives.
+`agentlog show <session-id> --unredacted` re-renders un-redacted from local cache. Local-only — never works on remote/team archives.
 
 **Honesty about limits:** pattern-based redaction catches credentials. It does not catch personal/sensitive content (medical, legal, financial conversations). This matters especially for web chat imports, which is why those default to local-only storage.
 
@@ -370,9 +372,9 @@ Web chats are excluded unless `include_web_chats=true`.
 **Adding to agents:**
 
 ```
-agentlog recall add-to claude    # writes ~/.claude/mcp.json, /commands/recall.md, and /skills/agentlog-recall/SKILL.md
-agentlog recall add-to cursor    # writes ~/.cursor/mcp.json
-agentlog recall add-to codex     # writes ~/.codex/config.toml
+agentlog integrations add-to claude    # writes ~/.claude/mcp.json, /commands/recall.md, and /skills/agentlog-recall/SKILL.md
+agentlog integrations add-to cursor    # writes ~/.cursor/mcp.json
+agentlog integrations add-to codex     # writes ~/.codex/config.toml
 ```
 
 Generated recall commands and skills should let the agent choose the first
@@ -389,13 +391,13 @@ fragments.
 
 ## History viewer
 
-v0 ships a dependency-free local viewer behind `agentlog history --web`. It lists sessions in a repo tree sorted by last updated time, pages large folders with a load-more control, searches the same event-first recall index, filters by repo/provider/date, and opens full conversations through the CLI API. The static viewer follows shadcn/ui-style tokens and compact button/input/select/sidebar patterns without requiring a frontend build step. Stable `path:<hash>` keys remain valid archive identifiers for folders without git identity, but the viewer displays the local folder path. The transcript pane defaults to readable chat bubbles for user, assistant, system, and tool messages, with a markdown toggle for the canonical archive file. Tool rendering reads canonical events or normalized metadata first, uses category/icon/target fields for consistent Bash/edit/read/search/web/task/skill/MCP cards, and uses raw text patterns only for legacy archives.
+v0 ships a dependency-free local viewer behind `agentlog web`. It lists sessions in a repo tree sorted by last updated time, pages large folders with a load-more control, searches the same event-first recall index, filters by repo/provider/date, and opens full conversations through the CLI API. The static viewer follows shadcn/ui-style tokens and compact button/input/select/sidebar patterns without requiring a frontend build step. Stable `path:<hash>` keys remain valid archive identifiers for folders without git identity, but the viewer displays the local folder path. The transcript pane defaults to readable chat bubbles for user, assistant, system, and tool messages, with a markdown toggle for the canonical archive file. Tool rendering reads canonical events or normalized metadata first, uses category/icon/target fields for consistent Bash/edit/read/search/web/task/skill/MCP cards, and uses raw text patterns only for legacy archives.
 
 **Commands:**
 
 ```
 agentlog history                          # native app pointed at archive
-agentlog history --web                    # web UI on localhost:7824
+agentlog web                              # web UI on localhost:7824
 agentlog history "query" --provider codex-cli
 agentlog history --repo github.com/org/repo
 agentlog history --since 7d
@@ -416,19 +418,22 @@ Complete user-facing commands:
 ```
 # Setup and lifecycle
 agentlog init [--storage local|r2|s3|custom] [--remote URL]
-agentlog start [--foreground]
-agentlog stop
+agentlog watcher start [--foreground]
+agentlog watcher stop
+agentlog watcher logs [--follow]
+agentlog watcher login <enable|disable|status>
 agentlog status
-agentlog logs [--follow]
-agentlog config <get|set> <key> [value]
-agentlog migrate --to <backend>
-agentlog sync [--endpoint <url>] [--bucket <name>] [--access-key-id <id>] [--secret-access-key <key>] [--prefix agentlog]
-agentlog autostart <enable|disable|status>
+agentlog config <show|path|get|set|setup|sources> [args]
+agentlog sync configure
+agentlog sync [--endpoint <url>] [--bucket <name>] [--access-key-id <id>] [--secret-access-key <key>] [--prefix agentlog] [--yes|--dry-run]
+agentlog sync snapshot [--name <label>] [--yes|--dry-run]
+agentlog sync replace
+agentlog sync wipe [--scope device|snapshot|snapshots|prefix|bucket] [--snapshot-name <name>] [--yes|--dry-run]
 agentlog doctor [--json]
 agentlog uninstall [--keep-data]
 
 # Capture management
-agentlog reveal <session-id>
+agentlog show <session-id> --unredacted
 agentlog redact reapply
 agentlog index <pause|resume|status>
 
@@ -439,16 +444,18 @@ agentlog import chatgpt --file <path> [--scope local|team]
 agentlog import status
 
 # Recall
-agentlog recall start
-agentlog recall add-to <codex|claude|gemini|antigravity|cursor>
-agentlog recall reindex
+agentlog mcp serve
+agentlog integrations add-to <codex|claude|gemini|antigravity|cursor>
+agentlog integrations recall [target]
+agentlog index rebuild
 
 # Viewing
-agentlog history [query] [--web] [--repo <repo>] [--provider <provider>] [--since <duration>] [--include-web-chats]
+agentlog history [query] [--repo <repo>] [--provider <provider>] [--since <duration>] [--include-web-chats]
+agentlog web [--port <port>] [--no-open]
 agentlog show <session-id> [--json|--path|--open]
 
 # Team mode
-agentlog server
+agentlog mcp serve
 ```
 
 ## Setup flows
@@ -457,11 +464,11 @@ agentlog server
 
 ```
 brew install agentlog
-agentlog init                    # picks local storage, prompts auto-start
-                                 # → "Start agentlog automatically at login? [Y/n]"
+agentlog init                    # picks local storage, prompts for the login watcher
+                                 # → "Start watcher at login"
                                  # → scans for existing history
                                  # → "Import last 30 days? [Y/n]"
-                                 # → writes launch agent, starts supervisor
+                                 # → if checked, writes launch agent and starts supervisor
                                  # → import runs in background
 ```
 
@@ -510,7 +517,7 @@ Companion deployment modules: `agentlog/deploy-aws` (ECS + ALB + S3), `agentlog/
 Codified because they shape every other decision:
 
 1. **No phone-home telemetry.** agentlog itself ships zero usage analytics anywhere. Any future opt-in metrics live on a separate channel.
-2. **No filesystem scanning.** Only specific known paths: `~/.codex/`, `~/.claude/`, `~/.gemini/`, `~/.local/share/devin/cli/sessions.db`, Cursor's storage, `~/.agentlog/`, plus user-specified import file paths. Windsurf paths are excluded while Cascade transcripts remain encrypted.
+2. **No filesystem scanning.** Only specific known paths: `~/.codex/`, `~/.claude/`, `~/.gemini/`, `~/.local/share/devin/cli/sessions.db`, Cursor's storage, `~/.agentlog/`, plus user-specified import file paths. Windsurf encrypted cache paths are excluded while Cascade transcripts remain encrypted; user-selected downloaded trajectory Markdown files are allowed.
 3. **No process inspection beyond `pgrep`.** We check whether Cursor is running. We don't introspect what it's doing.
 4. **Redaction at ingest, not query.** Pattern-matching credentials never land in storage in the first place.
 5. **Reveal is local-only.** Un-redacted content is never reconstructible from team/remote archives.

package/bin/agentlog-recall.js

@@ -1,5 +1,7 @@
 #!/usr/bin/env node
 
+process.title = "agentlog-recall";
+
 const { runMcpServer } = require("../src/mcp");
 
 runMcpServer().catch((error) => {

package/bin/agentlog.js

@@ -1,5 +1,7 @@
 #!/usr/bin/env node
 
+process.title = agentlogProcessTitle(process.argv.slice(2));
+
 const { main } = require("../src/cli");
 
 main(process.argv.slice(2)).catch((error) => {
@@ -12,3 +14,13 @@ function formatCliError(error) {
   const message = error?.message || String(error);
   return /^Error:/i.test(message) ? message : `Error: ${message}`;
 }
+
+function agentlogProcessTitle(argv) {
+  const command = argv[0] || "cli";
+  if (command === "start" && argv.includes("--foreground")) return "agentlog-supervisor";
+  if (command === "import") return "agentlog-import";
+  if (command === "index") return "agentlog-index";
+  if (command === "sync") return "agentlog-sync";
+  if (command === "recall" || command === "server") return "agentlog-recall";
+  return `agentlog-${String(command).replace(/[^A-Za-z0-9._-]+/g, "-").slice(0, 32) || "cli"}`;
+}