[cleak]

Part of the interview-challenge overhaul (S1-R: schema delta + schema upgrade + gitignore coverage).

Summary

Default schema delta (2026-10-interview-edges), additive only, in schema/default-schema.toml:

• epic: has_risk -> risk, has_question -> open_question
• decision: has_risk -> risk
• risk: risk_for -> epic, risk_for -> decision
• open_question: question_for -> epic

The reverse pairs (has_risk/risk_for, has_question/question_for) already existed in [edge_types]. tempyr init still embeds the file via include_str!. On epic roots, MissingRisk and UnresolvedQuestion gaps can now come up through the existing classify_edge_gap. No tests broke and gap logic is unchanged.

tempyr schema upgrade [--dry-run | --apply] (new crates/tempyr-cli/src/commands/schema_cmd.rs):

• Deltas are named, versioned entries defined in code (SCHEMA_DELTAS). Each one lists (node_type, edge_type, target) additions and the [edge_types] reverse entries to add if they are missing.
• Edits go through toml_edit (now a workspace dep; 0.22 is already in the lockfile through toml), so comments and layout are kept. New entries reuse the indentation of the existing allowed_edges array: inline single-line arrays, multi-line arrays and [[...allowed_edges]] arrays-of-tables all work. CRLF files stay CRLF.
• Additions are keyed on the triple, so ones already present are skipped. Edges are skipped and reported when the node type, target type or edge type is missing from the user's schema.
• Applied delta names are recorded in .tempyr/config.toml as [schema] applied_deltas = [...] (also through toml_edit). A delta that is already recorded is never applied again, so an edge the user removes stays removed. The embedding and journal config loaders only read their own sections, and a test checks that the rewritten config still loads through JournalConfig.
• Without --apply the command is a dry run: it lists the additions and writes nothing. --apply writes the files. --json is supported.
• After upgrading, the schema is re-parsed with Schema and every added edge is checked with validate_edge.
• tempyr init records every known delta as applied, because the default schema already contains them.

.gitignore coverage: interview archives may contain raw user answers. .tempyr/sessions/ and .tempyr/spar/ are now added to .gitignore by tempyr init and by tempyr schema upgrade --apply, and tempyr doctor warns when either is missing. There are no duplicate entries: equivalent patterns and parent patterns such as /.tempyr/sessions or .tempyr/ count as covered. The .env.local handling in init now uses the same helper (commands/gitignore.rs).

Existing projects get the new edges with tempyr schema upgrade --apply.

Test plan

• cargo fmt --check
• cargo clippy --all-targets (no warnings)
• cargo test (whole workspace)
• Unit tests (schema_cmd):
  • adds the edges to a pre-delta schema, and the output is byte-identical to the shipped default
  • running twice changes nothing the second time
  • comments and other config sections are preserved
  • a recorded delta is skipped
  • edges already present are not duplicated
  • missing node types are reported
  • array-of-tables layout works
  • CRLF line endings are preserved
  • the init block records every delta
• Unit tests (gitignore): appends once, adds only uncovered entries, creates the file, recognises equivalent and parent patterns
• Integration tests:
  • init writes the ignores and the applied_deltas, and schema upgrade then reports up to date
  • dry run lists the changes and writes nothing; doctor warns
  • --apply writes and leaves a valid schema, validate passes, and the doctor warning goes away
  • a recorded delta is not re-applied after the user removes an edge
  • --apply --dry-run is rejected
• Also tested every schema_cmd and integration test above against a CRLF copy of default-schema.toml

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

• Run on-demand review

_{This review includes 11 billable files and costs up to $2.75.}

• Ask an admin to make reviews automatic

Open in CodeRabbit

Reviews can continue after your included limit without a manual trigger. An admin must approve usage-based billing.

Or wait 57 minutes for your next included review.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available. Your 73 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: ASSERTIVE
• Plan: Essentials
• Run ID: e8b3407a-45be-4742-a014-9f877d158188

📥 Commits

Reviewing files that changed from the base of the PR and between 3c8da7a and 81d0075.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (11)

• Cargo.toml
• README.md
• crates/tempyr-cli/Cargo.toml
• crates/tempyr-cli/src/commands/doctor.rs
• crates/tempyr-cli/src/commands/gitignore.rs
• crates/tempyr-cli/src/commands/init.rs
• crates/tempyr-cli/src/commands/mod.rs
• crates/tempyr-cli/src/commands/schema_cmd.rs
• crates/tempyr-cli/src/main.rs
• crates/tempyr-cli/tests/integration.rs
• schema/default-schema.toml

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}