Release v0.5.4 · pleasedodisturb/chronoclaude

Zero-token /timestamps via scripts/timeline.sh. A shell wrapper that runs parse-transcript.py directly from your shell — no model turn, no tokens — and prints the session timeline (bash scripts/timeline.sh [count]). It locates the transcript with the same in-file cwd matching as /timestamps, so it works from any project path. Adapted from the timeline.sh in the s-a-s-k-i-a/claude-code-timestamps upstream (MIT); see THIRD-PARTY-LICENSES.md.

/timestamps failed to find the transcript for project paths containing ., spaces, or other rewritten characters. The slash command reconstructed the ~/.claude/projects/<dir> name with pwd | sed 's|/|-|g', which only rewrites / — so it mismatched Claude Code's actual folder name whenever the working-directory path held a character Claude Code also rewrites (dots, spaces, …), reporting "No transcript found." parse-transcript.py now resolves the transcript by matching the working directory recorded inside each transcript via a new --cwd argument (the robust approach from the s-a-s-k-i-a/claude-code-timestamps upstream), and /timestamps passes --cwd instead of guessing the folder name. The explicit-path CLI form is unchanged.

Install

/plugin marketplace add pleasedodisturb/chronoclaude
/plugin install chronoclaude@chronoclaude