v0.5.0: Journal Snapshot Command

New Features

jot stats — LLM-first journal snapshot data

Pre-computed journal snapshot data as JSON, designed for LLM consumption. Five sections, all optional:

jot stats                              # default: summary + overdue + blocked
jot stats --all                        # all five sections
jot stats --section=health             # specific section
jot stats --tags=myproject             # scoped to a tag
jot stats --type=task --status=open    # scoped by type/status

Sections:

Section	Description
summary	Aggregate counts by type, status, priority, tag count
overdue	Active entries past their due date
blocked	Entries with status=blocked, sorted by staleness
health	Per-tag project health: open/done/blocked/overdue/stale counts
recent	Entries modified in the last 7 days

Key design decisions:

• Always outputs JSON (LLM-first — intelligence stays in the LLM layer)
• Single filesystem walk regardless of how many sections are requested
• Reuses existing store.Filter flags for scoping (--tags, --type, --status, --since, etc.)
• --stale-days flag controls the staleness threshold for the health section (default: 30)

Other improvements

• EntrySummary now includes blocked_by field (backward-compatible via omitempty)

v0.4.0: Tag Management Command

New Features

jot tag command — first-class tag management

Three subcommands for managing entry tags without editing YAML frontmatter:

jot tag add <slug> <tag1,tag2,...>   # Add tags (additive, deduplicates)
jot tag rm <slug> <tag1,tag2,...>    # Remove tags (silent if missing)
jot tag set <slug> <tag1,tag2,...>   # Replace all tags

Batch mode via --stdin — reads slugs from stdin for pipeline composability:

# Tag all open tasks with sprint-5
jot list --type=task --status=open --json | jq -r '.slug' | jot tag add --stdin sprint-5

# Remove a tag from all entries that have it
jot list --tags=old-name --json | jq -r '.slug' | jot tag rm --stdin old-name

Other improvements

• parseTags() now deduplicates input tags
• Updated CLAUDE.md with current architecture documentation

v0.3.0

What's Changed

Bug Fixes

• Fix nil-pointer panic in new — os.Stat error on temp file was silently ignored; dereference of nil FileInfo caused panic
• Fix nil-pointer panic in show — os.Stdout.Stat() error ignored; now defaults to terminal rendering when stat fails
• Fix duplicate --force flag in rm — was a cobra-conflicting alias for --yes; removed
• Defensive yaml.Marshal handling in entry.ToMarkdown — writes error comment instead of silently producing empty YAML

Enhancements

• done works on any entry type — no longer restricted to tasks; ideas, notes, plans can all be marked done
• search gains --status and --priority filters — combine with text search for targeted results
• lint supports partial slug matching — uses ResolveSlug like other commands

Breaking Changes

• --tag renamed to --tags in list and search commands for consistency

Code Quality

• Fix variable shadowing in lint.go error loop
• Fix gofmt formatting in config_test.go
• Add tests for done-on-non-tasks, search filters, lint partial slugs

v0.2.0

Bug Fixes

• Fix infinite loop in highlightMatch — empty search query caused strings.Index to match at position 0 indefinitely
• Fix timezone mismatch in filterByDue — due dates parsed as UTC were incorrectly compared against local-timezone "today", causing off-by-one-day errors
• Fix ParseDate silent failure — invalid date strings now return an error instead of zero time with nil error

Code Quality

• Replace manual bubble sorts with sort.Slice in list and tasks commands
• Extract shared truncateSlug/truncateTitle helpers to eliminate duplication
• Remove unused mustGetStore function and dead imports
• Replace manual status validation loop with entry.Contains()
• Define named entryImporter interface to replace 3 identical inline declarations
• Simplify path manipulation in slug.go with strings.TrimSuffix
• Clean up parsePositiveInt to idiomatic (int, error) return signature

Tests

• Add command-level unit tests and integration tests (1,300+ lines)
• All 78 tests passing