Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/josephyaduvanshi/claude-history-manager/llms.txt

Use this file to discover all available pages before exploring further.

Chronicle’s search bar combines instant title search with a set of slash-command filters that let you narrow sessions by date, tag, workspace, or transcript content. Filters compose left-to-right in a single query string — no separate filter UI to open. All results are scoped to the active provider; switching from Claude to Codex clears the results and re-runs the query against the new provider’s index. Press ⌘K to focus the search bar from anywhere in the app.

​
How parsing works

The query bar parses your input left-to-right, token by token. Slash-prefixed tokens (/today, /tag:bug, /in:rust) set structured filters. Everything else accumulates as free-form title text matched with a LIKE query against indexed session titles. Unknown slash tokens are treated as plain title text — a typo like /tday won’t silently disappear from your query.
The default result limit is 500 rows. The limit applies after all filters are evaluated.

​
Slash filter reference

Searches across the full transcript bodies of every session, not just titles. Uses an FTS5 virtual table built lazily on first use — the index is constructed the first time you type /full: and cached for subsequent searches.The /full: token is greedy: everything that follows on the same line is treated as the search query, including spaces. You do not need quotes for multi-word queries.
/full: EXC_BAD_ACCESS
/full: async let task cancelled
Filters to sessions whose last_modified_at timestamp falls within the current calendar day (local time).
/today
/today refactor
Both spellings are equivalent. Filters to sessions modified within the last 7 days.
/this-week
/thisweek auth
All three spellings are equivalent. Filters to sessions modified within the last 30 days.
/last30days
/last-30d performance
/last30d
Filters to sessions carrying a specific tag. Use double quotes for tag names that contain spaces.
/tag:bug
/tag:client-work
/tag:"client work"
You can apply multiple /tag: filters in the same query; only sessions carrying all specified tags are returned.
/tag:bug /tag:regression
Filters to sessions in a specific workspace, matched against the workspace’s display name or decoded filesystem path. Use double quotes for workspace names or paths that contain spaces.
/in:rust
/in:flutter
/in:"my project"
/in:"/Users/you/Code/flutter/myapp"
Like /tag:, multiple /in: filters are supported and ANDed together.

​
Combining filters

Filters compose in the order you type them. Any non-slash tokens between or after slash tokens become the title search text. The query is always evaluated as a conjunction — all active filters must match.
QueryWhat it means
/in:rust /tag:bug panicRust workspaces, tagged bug, title contains “panic”
/today /full: connection refusedSessions from today whose transcript mentions “connection refused”
/this-week /tag:client-work invoiceThis week’s sessions tagged client-work, title contains “invoice”
/last30days /in:flutter /tag:crashFlutter sessions in the last 30 days tagged crash
The title text portion of the query (anything that isn’t a slash token) uses a case-insensitive LIKE match. You don’t need wildcards — panic matches “add panic handler” and “panic in thread”.

​
Full-text search details

Full-text search with /full: queries an FTS5 virtual table over the transcript bodies of all sessions. This table is built lazily the first time you use /full: — expect a short delay on that first search while the index is constructed. Subsequent /full: queries use the cached index.
FTS5 indexing is per-provider. Switching providers discards the current result set; the FTS index for the new provider is built on demand when you next use /full:.

Build docs developers (and LLMs) love

Get started for freeTalk to us