feat(phase 4): branch summary preservation in /tree (pi-style metadata) - #72
Merged
Conversation
Phase 4 of the 6-phase plan. Reference pattern: pi's
`packages/coding-agent/src/core/branch-summarization.ts` —
sibling branches dropped during compaction get preserved as
structured summary records so the user can see what was lost.
dirge implements the metadata-only variant (no LLM-generated
summary call yet) — the schema is forward-compatible if Phase 4b
adds LLM summaries later.
## Problem
Phase 2 added sibling-branch pruning with a chat notification:
"discarded N forked branches". Useful in the moment but the
information was lost forever. Users with branched sessions
couldn't see WHICH branches were dropped or any preview of what
was in them.
## Fix
New `BranchSummary` struct in `src/session/mod.rs`:
```rust
pub struct BranchSummary {
pub root_id: CompactString, // top of the pruned subtree
pub parent_id: CompactString, // still-present parent (may be dropped too)
pub message_count: usize,
pub preview: String, // "[label] first 80 chars..."
pub created_at: String, // RFC3339
}
```
`Session::branch_summaries: Vec<BranchSummary>` with
`#[serde(default)]` for back-compat. Cleared by `reset_to_new`
(else fresh sessions would inherit phantom records).
### Capture during compress + rewind
Before pruning, walk the `to_prune` set and identify SUBTREE
ROOTS — nodes in `to_prune` whose direct parent was in
`dropped_set` (the closest dropped-active-path ancestor). For
each root, build a `BranchSummary`:
- `root_id` and `parent_id` for correlation
- Walk descendants to count the subtree size
- Preview = optional label (`[explore-alt]`) + first 80 chars of
the root's message content
Identical algorithm in `Session::compress_reporting` and
`ui::mod.rs::rewind_session`.
### Surface in `/tree`
`render_tree` appends a "Summarized branches" section after the
linear tree render:
```
* abc12345 user "the active path…"
abc12346 asst "the next message…"
Summarized branches (1): pruned during compress/rewind
└─ parent abc12345 · 2 msgs · [explore-alt] let me try a different…
```
Users browsing `/tree` after a compress can now see exactly
which branches were preserved and which were lost.
## Tests
2 new session tests, written failing first:
- `compress_records_branch_summary_for_pruned_siblings`:
builds a session with u1 → [a1, u2, a2] linear + a sibling
subtree (sib_alpha labeled "explore-alt" → sib_beta) rooted
at u1, compresses past u1, asserts the resulting
`branch_summaries` has 1 entry with the right parent_id (u1),
message_count (2), and preview containing "explore-alt" or
the alpha sibling's content.
- `reset_to_new_clears_branch_summaries`: `/clear`-equivalent
wipes the summary list along with everything else.
Total: 646 pass (was 644), 0 fail, all build profiles clean.
## Not in Phase 4 (deferred to optional Phase 4b)
- LLM-generated summary content. Today the preview is the
truncated content + label; richer one-sentence summaries
would require routing the LLM client into `Session::compress`
(currently sync) — bigger plumbing change. The
`BranchSummary` schema can accommodate a future `summary:
Option<String>` field without breaking back-compat.
- Restoration as agent context when the user navigates "back"
to a summarized branch. Today summaries are display-only
(`/tree`). Navigation-time injection would need a new tree
op (e.g. `harness/peek-summary` or a system message
injection) — out of scope for the metadata-only MVP.
## Test plan
- [x] `cargo test --features plugin` -> 646 pass.
- [x] `cargo build --all-features` -> compiles.
- [x] `cargo build --no-default-features` -> compiles.
## Up next: Phase 5 (per user request, skip Phase 6)
`/allow list | remove <idx> | add <tool> <pattern>` CRUD slash
command. User asked to skip Phase 6 (cost tracking).
yogthos
added a commit
that referenced
this pull request
May 21, 2026
) Phase 5 of the plan. No clean equivalent in opencode or pi — both manage allowlist entries only via the interactive permission prompt's "allow always" answer. dirge adds explicit CRUD so users can inspect, manually add, drop one, or clear all entries without editing the session JSON. ## Commands - `/allow` or `/allow list` — show numbered entries - `/allow add <tool> <pattern>` — add a (tool, pattern) entry, e.g. `/allow add bash 'cargo *'` - `/allow remove <idx>` — drop one entry by 0-based index from the `list` output - `/allow clear` — drop all entries The `add` form preserves pattern strings containing spaces by re-deriving the args from the raw `text` slice (rather than relying on the SmallVec `parts[2]` which would get truncated at the first whitespace). Same approach `/cd` uses for paths. ## State sync Both surfaces stay aligned: - `PermissionChecker::session_allowlist` (in-memory, used by `check` / `check_path`) — updated via the existing `add_session_allowlist` and the two new `remove_session_allowlist_at(idx)` + `clear_session_allowlist`. - `Session::permission_allowlist` (persisted to JSON, restored on resume via `load_session_allowlist`) — mirrored manually in the slash handler so save/load round-trips show the user-edited list. Without the session-side mirror, a `/allow add foo bar` would stay in memory for this run but vanish on `-c` resume. Dedup at the session level too so `save()` writes a clean list. ## Tests 3 new permission tests, written failing first: - `remove_session_allowlist_at_returns_removed_entry`: remove(1) on [bash:cargo*, bash:git*, read:/tmp/*] yields Some(("bash","git *")) and the surviving entries shift down. - `remove_session_allowlist_at_out_of_range_returns_none`: remove(99) returns None without panicking; existing entries intact. - `clear_session_allowlist_empties_the_list`: clear() drops all entries. The slash handler itself is not unit-tested (requires the full UI loop fixtures); manual smoke test: ``` $ dirge --restrictive > bash 'echo hi' > (a) allow always > /allow list → [0] bash echo * > /allow add read /tmp/* > /allow list → [0] bash echo * [1] read /tmp/* > /allow remove 0 > /allow list → [0] read /tmp/* > /allow clear > /allow list → empty ``` ## Docs `/help` text gets four new lines covering each subcommand. ## Test plan - [x] `cargo test --features plugin` -> 649 pass (was 646). - [x] `cargo build --all-features` -> compiles. - [x] `cargo build --no-default-features` -> compiles. ## Plan status | Phase | PR | Done | |---|---|---| | 1 | #69 | first-wins block + docs | | 2 | #70 | sibling-branch pruning + notification | | 3 | #71 | structured tool-call persistence | | 4 | #72 | branch summary metadata in /tree | | 5 | this | /allow CRUD | | 6 | — | skipped per user request (cost tracking) | 5 of 6 phases complete; Phase 6 (cost tracking) skipped. Co-authored-by: Yogthos <yogthos@gmail.com>
allen-munsch
pushed a commit
to allen-munsch/dirge
that referenced
this pull request
Jun 3, 2026
…a) (dirge-code#72) Phase 4 of the 6-phase plan. Reference pattern: pi's `packages/coding-agent/src/core/branch-summarization.ts` — sibling branches dropped during compaction get preserved as structured summary records so the user can see what was lost. dirge implements the metadata-only variant (no LLM-generated summary call yet) — the schema is forward-compatible if Phase 4b adds LLM summaries later. ## Problem Phase 2 added sibling-branch pruning with a chat notification: "discarded N forked branches". Useful in the moment but the information was lost forever. Users with branched sessions couldn't see WHICH branches were dropped or any preview of what was in them. ## Fix New `BranchSummary` struct in `src/session/mod.rs`: ```rust pub struct BranchSummary { pub root_id: CompactString, // top of the pruned subtree pub parent_id: CompactString, // still-present parent (may be dropped too) pub message_count: usize, pub preview: String, // "[label] first 80 chars..." pub created_at: String, // RFC3339 } ``` `Session::branch_summaries: Vec<BranchSummary>` with `#[serde(default)]` for back-compat. Cleared by `reset_to_new` (else fresh sessions would inherit phantom records). ### Capture during compress + rewind Before pruning, walk the `to_prune` set and identify SUBTREE ROOTS — nodes in `to_prune` whose direct parent was in `dropped_set` (the closest dropped-active-path ancestor). For each root, build a `BranchSummary`: - `root_id` and `parent_id` for correlation - Walk descendants to count the subtree size - Preview = optional label (`[explore-alt]`) + first 80 chars of the root's message content Identical algorithm in `Session::compress_reporting` and `ui::mod.rs::rewind_session`. ### Surface in `/tree` `render_tree` appends a "Summarized branches" section after the linear tree render: ``` * abc12345 user "the active path…" abc12346 asst "the next message…" Summarized branches (1): pruned during compress/rewind └─ parent abc12345 · 2 msgs · [explore-alt] let me try a different… ``` Users browsing `/tree` after a compress can now see exactly which branches were preserved and which were lost. ## Tests 2 new session tests, written failing first: - `compress_records_branch_summary_for_pruned_siblings`: builds a session with u1 → [a1, u2, a2] linear + a sibling subtree (sib_alpha labeled "explore-alt" → sib_beta) rooted at u1, compresses past u1, asserts the resulting `branch_summaries` has 1 entry with the right parent_id (u1), message_count (2), and preview containing "explore-alt" or the alpha sibling's content. - `reset_to_new_clears_branch_summaries`: `/clear`-equivalent wipes the summary list along with everything else. Total: 646 pass (was 644), 0 fail, all build profiles clean. ## Not in Phase 4 (deferred to optional Phase 4b) - LLM-generated summary content. Today the preview is the truncated content + label; richer one-sentence summaries would require routing the LLM client into `Session::compress` (currently sync) — bigger plumbing change. The `BranchSummary` schema can accommodate a future `summary: Option<String>` field without breaking back-compat. - Restoration as agent context when the user navigates "back" to a summarized branch. Today summaries are display-only (`/tree`). Navigation-time injection would need a new tree op (e.g. `harness/peek-summary` or a system message injection) — out of scope for the metadata-only MVP. ## Test plan - [x] `cargo test --features plugin` -> 646 pass. - [x] `cargo build --all-features` -> compiles. - [x] `cargo build --no-default-features` -> compiles. ## Up next: Phase 5 (per user request, skip Phase 6) `/allow list | remove <idx> | add <tool> <pattern>` CRUD slash command. User asked to skip Phase 6 (cost tracking). Co-authored-by: Yogthos <yogthos@gmail.com>
allen-munsch
pushed a commit
to allen-munsch/dirge
that referenced
this pull request
Jun 3, 2026
…irge-code#73) Phase 5 of the plan. No clean equivalent in opencode or pi — both manage allowlist entries only via the interactive permission prompt's "allow always" answer. dirge adds explicit CRUD so users can inspect, manually add, drop one, or clear all entries without editing the session JSON. ## Commands - `/allow` or `/allow list` — show numbered entries - `/allow add <tool> <pattern>` — add a (tool, pattern) entry, e.g. `/allow add bash 'cargo *'` - `/allow remove <idx>` — drop one entry by 0-based index from the `list` output - `/allow clear` — drop all entries The `add` form preserves pattern strings containing spaces by re-deriving the args from the raw `text` slice (rather than relying on the SmallVec `parts[2]` which would get truncated at the first whitespace). Same approach `/cd` uses for paths. ## State sync Both surfaces stay aligned: - `PermissionChecker::session_allowlist` (in-memory, used by `check` / `check_path`) — updated via the existing `add_session_allowlist` and the two new `remove_session_allowlist_at(idx)` + `clear_session_allowlist`. - `Session::permission_allowlist` (persisted to JSON, restored on resume via `load_session_allowlist`) — mirrored manually in the slash handler so save/load round-trips show the user-edited list. Without the session-side mirror, a `/allow add foo bar` would stay in memory for this run but vanish on `-c` resume. Dedup at the session level too so `save()` writes a clean list. ## Tests 3 new permission tests, written failing first: - `remove_session_allowlist_at_returns_removed_entry`: remove(1) on [bash:cargo*, bash:git*, read:/tmp/*] yields Some(("bash","git *")) and the surviving entries shift down. - `remove_session_allowlist_at_out_of_range_returns_none`: remove(99) returns None without panicking; existing entries intact. - `clear_session_allowlist_empties_the_list`: clear() drops all entries. The slash handler itself is not unit-tested (requires the full UI loop fixtures); manual smoke test: ``` $ dirge --restrictive > bash 'echo hi' > (a) allow always > /allow list → [0] bash echo * > /allow add read /tmp/* > /allow list → [0] bash echo * [1] read /tmp/* > /allow remove 0 > /allow list → [0] read /tmp/* > /allow clear > /allow list → empty ``` ## Docs `/help` text gets four new lines covering each subcommand. ## Test plan - [x] `cargo test --features plugin` -> 649 pass (was 646). - [x] `cargo build --all-features` -> compiles. - [x] `cargo build --no-default-features` -> compiles. ## Plan status | Phase | PR | Done | |---|---|---| | 1 | dirge-code#69 | first-wins block + docs | | 2 | dirge-code#70 | sibling-branch pruning + notification | | 3 | dirge-code#71 | structured tool-call persistence | | 4 | dirge-code#72 | branch summary metadata in /tree | | 5 | this | /allow CRUD | | 6 | — | skipped per user request (cost tracking) | 5 of 6 phases complete; Phase 6 (cost tracking) skipped. Co-authored-by: Yogthos <yogthos@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Phase 4 of the 6-phase plan. Pruned sibling branches now get
BranchSummaryrecords (root id, parent id, count, preview) preserved on the session./treeshows a 'Summarized branches' section after the linear tree so users can see what was discarded. Metadata-only MVP; richer LLM-summary variant is Phase 4b. 2 new tests, 646 pass.