chore(deps): bump @os-eco/mulch-cli from 0.6.5 to 0.7.0

[dependabot]

Bumps @os-eco/mulch-cli from 0.6.5 to 0.7.0.

Release notes

Sourced from @​os-eco/mulch-cli's releases.

v0.7.0

Added

Global --format Flag

• Global --format <markdown|compact|xml|plain> flag that routes record-rendering commands (ml prime, ml query, ml search) through the selected formatter. xml is Claude-optimized, plain is Codex-optimized, compact emits one-liners (the default for ml prime), markdown emits the full sectioned layout. Per-command --format (including ml query --format ids and ml search --format ids) still wins over the global flag. ml prime --compact and ml prime --full are kept as aliases for --format compact and --format markdown respectively.
• ml query --format xml|plain and ml search --format xml|plain — the four formatters are now reachable from any record-rendering command, not just ml prime.

Manifest Mode for Monoliths

• ml prime --manifest emits a quick reference + per-domain index (with per-record-type counts and governance status) instead of full records — designed for monolith projects where dumping every record across every domain wastes agent context
• Optional prime.default_mode config knob in .mulch/mulch.config.yaml: set to manifest so plain ml prime (with no scoping args) emits the index by default; ml prime <domain> and ml prime --files <path> keep loading full records for the requested scope
• --full flag forces full output even when config says manifest
• --manifest combined with any scoping argument (<domain>, --domain, --exclude-domain, --context, --files) is a hard error with a usage hint
• ml prime --manifest --json emits a structured { type: "manifest", quick_reference, domains[] } payload with per-type counts and per-domain health status

Search Confirmation-Frequency Boost

• ml search <query> now applies a confirmation-frequency boost to BM25 scores by default: records with successful outcomes float above unconfirmed records at the same relevance. Boost factor is 1 + 0.1 * confirmation_score per record (records with no outcomes are unaffected). Activates the previously-unused applyConfirmationBoost helper.
• Optional search.boost_factor knob in .mulch/mulch.config.yaml to tune or disable the boost (0 = pure BM25)
• ml search --no-boost flag as a per-call escape hatch back to pure BM25 ordering. --sort-by-score is unchanged and continues to work as a confirmation-only post-sort.

Init Scaffolding

• ml init now writes .mulch/mulch.config.yaml from a templated string with a header comment and a commented-out optional-knobs section (currently prime.default_mode) so users can discover settings without reading the source. The body is generated via yaml.dump(DEFAULT_CONFIG) so required-field values can't drift; subsequent writeConfig() calls still round-trip through the YAML serializer and strip comments — by design, the scaffold lives only at init time.

Tooling

• scripts/version-bump.ts — atomically bumps package.json and src/cli.ts in lockstep; wired into bun run version:bump <major|minor|patch>. (The script existed in package.json since 2026-03-05 but was never committed; fresh clones used to fail to bump.)

Changed

• ml onboard snippet rewritten to cover 0.6.4 agent workflow: multi-tracker evidence (--evidence-seeds/--evidence-gh/--evidence-linear), git auto-context for commit + files, --relates-to, outcome merge on upserts, retry hints on validation failures, ml doctor --fix for broken file anchors, and worktree-safe storage; now also mentions manifest mode for monolith discovery
• ONBOARD_VERSION bumped to 3 so existing v:1 and v:2 installs are detected as outdated and migrated on the next ml onboard
• MULCH_README (rendered to .mulch/README.md on ml init) now documents the optional prime.default_mode knob
• formatMcpOutput / McpDomain renamed to formatJsonOutput / JsonDomain to match reality (no MCP integration was ever wired up)
• TypeScript bumped from 5.9.3 to 6.0.3 (Dependabot #20); bun.lock regenerated to match

Removed

• ml prime --mcp flag (use --json instead — the two produced identical output and there was no MCP integration consuming the flag)

Testing

• 840 tests across 41 files, 1935 expect() calls (up from 811 / 41 / 1848 in 0.6.5)
• Expanded test/commands/prime.test.ts (+401 lines): manifest-mode output, scoping-conflict errors, prime.default_mode config resolution, --full override, JSON manifest payload shape
• Expanded test/commands/query.test.ts (+122 lines) and test/commands/search.test.ts (+140 lines): global --format resolution across xml/plain/markdown/compact, per-command --format precedence
• Expanded test/commands/init.test.ts (+26 lines): scaffolded YAML header comment and commented-out prime.default_mode block

Changelog

Sourced from @​os-eco/mulch-cli's changelog.

[0.7.0] - 2026-04-28

Added

Global --format Flag

• Global --format <markdown|compact|xml|plain> flag that routes record-rendering commands (ml prime, ml query, ml search) through the selected formatter. xml is Claude-optimized, plain is Codex-optimized, compact emits one-liners (the default for ml prime), markdown emits the full sectioned layout. Per-command --format (including ml query --format ids and ml search --format ids) still wins over the global flag. ml prime --compact and ml prime --full are kept as aliases for --format compact and --format markdown respectively.
• ml query --format xml|plain and ml search --format xml|plain — the four formatters are now reachable from any record-rendering command, not just ml prime.

Manifest Mode for Monoliths

• ml prime --manifest emits a quick reference + per-domain index (with per-record-type counts and governance status) instead of full records — designed for monolith projects where dumping every record across every domain wastes agent context
• Optional prime.default_mode config knob in .mulch/mulch.config.yaml: set to manifest so plain ml prime (with no scoping args) emits the index by default; ml prime <domain> and ml prime --files <path> keep loading full records for the requested scope
• --full flag forces full output even when config says manifest
• --manifest combined with any scoping argument (<domain>, --domain, --exclude-domain, --context, --files) is a hard error with a usage hint
• ml prime --manifest --json emits a structured { type: "manifest", quick_reference, domains[] } payload with per-type counts and per-domain health status

Search Confirmation-Frequency Boost

• ml search <query> now applies a confirmation-frequency boost to BM25 scores by default: records with successful outcomes float above unconfirmed records at the same relevance. Boost factor is 1 + 0.1 * confirmation_score per record (records with no outcomes are unaffected). Activates the previously-unused applyConfirmationBoost helper.
• Optional search.boost_factor knob in .mulch/mulch.config.yaml to tune or disable the boost (0 = pure BM25)
• ml search --no-boost flag as a per-call escape hatch back to pure BM25 ordering. --sort-by-score is unchanged and continues to work as a confirmation-only post-sort.

Init Scaffolding

• ml init now writes .mulch/mulch.config.yaml from a templated string with a header comment and a commented-out optional-knobs section (currently prime.default_mode) so users can discover settings without reading the source. The body is generated via yaml.dump(DEFAULT_CONFIG) so required-field values can't drift; subsequent writeConfig() calls still round-trip through the YAML serializer and strip comments — by design, the scaffold lives only at init time.

Tooling

• scripts/version-bump.ts — atomically bumps package.json and src/cli.ts in lockstep; wired into bun run version:bump <major|minor|patch>. (The script existed in package.json since 2026-03-05 but was never committed; fresh clones used to fail to bump.)

Changed

• ml onboard snippet rewritten to cover 0.6.4 agent workflow: multi-tracker evidence (--evidence-seeds/--evidence-gh/--evidence-linear), git auto-context for commit + files, --relates-to, outcome merge on upserts, retry hints on validation failures, ml doctor --fix for broken file anchors, and worktree-safe storage; now also mentions manifest mode for monolith discovery
• ONBOARD_VERSION bumped to 3 so existing v:1 and v:2 installs are detected as outdated and migrated on the next ml onboard
• MULCH_README (rendered to .mulch/README.md on ml init) now documents the optional prime.default_mode knob
• formatMcpOutput / McpDomain renamed to formatJsonOutput / JsonDomain to match reality (no MCP integration was ever wired up)
• TypeScript bumped from 5.9.3 to 6.0.3 (Dependabot #20); bun.lock regenerated to match

Removed

• ml prime --mcp flag (use --json instead — the two produced identical output and there was no MCP integration consuming the flag)

Testing

• 840 tests across 41 files, 1935 expect() calls (up from 811 / 41 / 1848 in 0.6.5)
• Expanded test/commands/prime.test.ts (+401 lines): manifest-mode output, scoping-conflict errors, prime.default_mode config resolution, --full override, JSON manifest payload shape
• Expanded test/commands/query.test.ts (+122 lines) and test/commands/search.test.ts (+140 lines): global --format resolution across xml/plain/markdown/compact, per-command --format precedence
• Expanded test/commands/init.test.ts (+26 lines): scaffolded YAML header comment and commented-out prime.default_mode block

Commits

• b4016a6 Prepare v0.7.0 release: --format flag, manifest mode, search confirmation boost
• 7be7dda feat(cli): add global --format flag for record output
• 8b15ee1 seeds: sync 2026-04-29
• 5346ad9 mulch: update expertise
• 14f6974 feat(init): scaffold config with commented-out optional knobs
• 1cf4b42 seeds: sync 2026-04-29
• 4010387 mulch: update expertise
• 8767fb3 chore: track scripts/version-bump.ts referenced by version:bump npm script
• f9349f3 feat(prime): manifest mode for monolith projects; drop vestigial --mcp flag
• 1e4422a seeds: sync 2026-04-29
• Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

[dependabot]

Superseded by #150.