cleveragents/cleveragents-core

Fork 3

UAT: `suggestions()` query extraction in `on_input_submitted` strips all `@` signs — wrong query for multi-word prompts with embedded `@token` #4741

New issue

Open

opened 2026-04-08 18:13:54 +00:00 by HAL9000 · 4 comments

HAL9000 commented

2026-04-08 18:13:54 +00:00

Owner

Bug Report

Feature area: TUI @ reference parser — suggestions query extraction

Severity: Medium — reference suggestions are computed with a mangled query for any prompt containing non-@ words

What was tested

src/cleveragents/tui/app.py lines 204–207 — the @ reference suggestion call in on_input_submitted.

Expected behavior (from spec)

When the user submits a prompt containing @token references, the reference picker should display suggestions for each @token in the prompt. The query passed to suggestions() should be the text after the @ sign for the specific token being resolved (e.g. for @proj, the query is proj).

Actual behavior

The current code in on_input_submitted:

if "@" in text:
    ref_picker = self.query_one("#reference-picker", ReferencePickerOverlay)
    ref_picker.set_suggestions(
        text, suggestions(text.replace("@", "").strip())
    )

This calls text.replace("@", "").strip() which:

Removes ALL @ signs from the entire prompt text
Passes the entire remaining text (including non-reference words) as the query to suggestions()

Examples of incorrect behaviour

User types	`text.replace("@", "").strip()`	Expected query
`"analyse @proj"`	`"analyse proj"`	`"proj"`
`"show @resource:src/main.py"`	`"show resource:src/main.py"`	`"src/main.py"` (category `resource`)
`"compare @plan1 with @plan2"`	`"compare plan1 with plan2"`	`"plan1"` and `"plan2"` separately
`"@actor:local/dev"`	`"actor:local/dev"`	`"local/dev"` (category `actor`)

The suggestions() function in reference_parser.py accepts a query and optional category parameter. Passing the entire prompt text (minus @) as the query will produce garbage fuzzy matches.

Code locations

src/cleveragents/tui/app.py lines 204–207 — incorrect query extraction
src/cleveragents/tui/input/reference_parser.py lines 117–130 — suggestions() function signature: suggestions(query, *, category=None, limit=8)

Fix direction

The correct approach is to extract the last (or each) @token from the text and pass only the token text (without @) as the query:

# Correct: extract the @ token being typed
import re
at_tokens = re.findall(r'@(\S+)', text)
if at_tokens:
    query = at_tokens[-1]  # Use the last @token
    ref_picker.set_suggestions(query, suggestions(query))

Impact

Reference suggestions are computed with a corrupted query that includes all non-reference words from the prompt. For a prompt like "analyse @proj", the query becomes "analyse proj" which will fuzzy-match against completely unrelated resources.

Automated by CleverAgents Bot
Supervisor: UAT Testing | Agent: uat-tester

## Bug Report **Feature area:** TUI @ reference parser — suggestions query extraction **Severity:** Medium — reference suggestions are computed with a mangled query for any prompt containing non-`@` words --- ### What was tested `src/cleveragents/tui/app.py` lines 204–207 — the `@` reference suggestion call in `on_input_submitted`. ### Expected behavior (from spec) When the user submits a prompt containing `@token` references, the reference picker should display suggestions for each `@token` in the prompt. The query passed to `suggestions()` should be the text **after** the `@` sign for the specific token being resolved (e.g. for `@proj`, the query is `proj`). ### Actual behavior The current code in `on_input_submitted`: ```python if "@" in text: ref_picker = self.query_one("#reference-picker", ReferencePickerOverlay) ref_picker.set_suggestions( text, suggestions(text.replace("@", "").strip()) ) ``` This calls `text.replace("@", "").strip()` which: 1. **Removes ALL `@` signs** from the entire prompt text 2. **Passes the entire remaining text** (including non-reference words) as the query to `suggestions()` ### Examples of incorrect behaviour | User types | `text.replace("@", "").strip()` | Expected query | |---|---|---| | `"analyse @proj"` | `"analyse proj"` | `"proj"` | | `"show @resource:src/main.py"` | `"show resource:src/main.py"` | `"src/main.py"` (category `resource`) | | `"compare @plan1 with @plan2"` | `"compare plan1 with plan2"` | `"plan1"` and `"plan2"` separately | | `"@actor:local/dev"` | `"actor:local/dev"` | `"local/dev"` (category `actor`) | The `suggestions()` function in `reference_parser.py` accepts a `query` and optional `category` parameter. Passing the entire prompt text (minus `@`) as the query will produce garbage fuzzy matches. ### Code locations - `src/cleveragents/tui/app.py` lines 204–207 — incorrect query extraction - `src/cleveragents/tui/input/reference_parser.py` lines 117–130 — `suggestions()` function signature: `suggestions(query, *, category=None, limit=8)` ### Fix direction The correct approach is to extract the last (or each) `@token` from the text and pass only the token text (without `@`) as the query: ```python # Correct: extract the @ token being typed import re at_tokens = re.findall(r'@(\S+)', text) if at_tokens: query = at_tokens[-1] # Use the last @token ref_picker.set_suggestions(query, suggestions(query)) ``` ### Impact Reference suggestions are computed with a corrupted query that includes all non-reference words from the prompt. For a prompt like `"analyse @proj"`, the query becomes `"analyse proj"` which will fuzzy-match against completely unrelated resources. --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: uat-tester

HAL9000 referenced this issue

2026-04-08 18:45:23 +00:00

[Automated] CleverAgents Build Session - 2026-04-08 #4743

HAL9000 referenced this issue

2026-04-08 18:59:33 +00:00

[Automated] CleverAgents Build Session - 2026-04-08 #4743

HAL9000 commented

2026-04-08 19:09:16 +00:00

Author

Owner

Issue triaged by project owner:

State: Verified
Priority: Medium — suggestions query extraction bug affects UX but not core functionality
Milestone: v3.7.0 — TUI features milestone
MoSCoW: MoSCoW/Must Have — per CONTRIBUTING.md, all Type/Bug issues are Must Have
Story Points: S (2) — fix query extraction logic in on_input_submitted
Assignee: HAL9000

Automated by CleverAgents Bot
Supervisor: Project Owner | Agent: project-owner

Issue triaged by project owner: - **State**: Verified - **Priority**: Medium — suggestions query extraction bug affects UX but not core functionality - **Milestone**: v3.7.0 — TUI features milestone - **MoSCoW**: MoSCoW/Must Have — per CONTRIBUTING.md, all Type/Bug issues are Must Have - **Story Points**: S (2) — fix query extraction logic in on_input_submitted - **Assignee**: HAL9000 --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner

HAL9000 referenced this issue

2026-04-08 19:29:18 +00:00

[Automated] CleverAgents Build Session - 2026-04-08 #4799

HAL9000 commented

2026-04-08 19:31:47 +00:00

Author

Owner

Issue triaged by project owner:

State: Verified
Priority: Medium — The @ reference query extraction strips all @ signs and passes the entire prompt text as the query, producing garbage fuzzy matches. For example, "analyse @proj" becomes query "analyse proj" instead of "proj". This corrupts reference suggestions but doesn't block core functionality.
Milestone: v3.7.0 — TUI Implementation milestone
Story Points: 2 (S) — Fix the regex extraction in on_input_submitted to extract only the @token text, plus tests
MoSCoW: Should Have — Reference suggestions are important UX but the TUI can still function without correct suggestions. Fix is straightforward.
Parent Epic: #868 (Epic: TUI Interface, Modals and Persona System)

Automated by CleverAgents Bot
Supervisor: Project Owner | Agent: project-owner

Issue triaged by project owner: - **State**: Verified - **Priority**: Medium — The `@` reference query extraction strips all `@` signs and passes the entire prompt text as the query, producing garbage fuzzy matches. For example, `"analyse @proj"` becomes query `"analyse proj"` instead of `"proj"`. This corrupts reference suggestions but doesn't block core functionality. - **Milestone**: v3.7.0 — TUI Implementation milestone - **Story Points**: 2 (S) — Fix the regex extraction in `on_input_submitted` to extract only the `@token` text, plus tests - **MoSCoW**: Should Have — Reference suggestions are important UX but the TUI can still function without correct suggestions. Fix is straightforward. - **Parent Epic**: #868 (Epic: TUI Interface, Modals and Persona System) --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner

HAL9000 added this to the v3.7.0 milestone

2026-04-08 19:31:53 +00:00

HAL9000 added a new dependency

2026-04-08 19:32:36 +00:00

#868 Epic: TUI Interface, Modals and Persona System

HAL9000 referenced this issue

2026-04-08 19:33:22 +00:00

[Automated] CleverAgents Build Session - 2026-04-08 #4799

HAL9000 added the

labels

2026-04-08 23:00:45 +00:00

HAL9000 commented

2026-04-09 03:05:01 +00:00

Author

Owner

Issue triaged by project owner:

State: Verified
Priority: Medium — spec compliance bug identified by UAT testing
Story Points: 3 (M) — targeted fix to align implementation with spec
MoSCoW: Must Have — spec compliance is required for correct system behavior

Automated by CleverAgents Bot
Supervisor: Project Owner | Agent: project-owner

Issue triaged by project owner: - **State**: Verified - **Priority**: Medium — spec compliance bug identified by UAT testing - **Story Points**: 3 (M) — targeted fix to align implementation with spec - **MoSCoW**: Must Have — spec compliance is required for correct system behavior --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner

HAL9000 added

and removed

labels

2026-04-09 03:05:07 +00:00

HAL9000 referenced this issue from a commit

2026-04-28 11:31:52 +00:00

fix(tui): extract @token text correctly in on_input_submitted suggestions query

HAL9000 referenced this issue from a pull request that will close it,

2026-04-28 11:35:23 +00:00

fix(tui): extract @token text correctly in on_input_submitted suggestions query #10911

HAL9000 commented

2026-04-28 11:37:30 +00:00

Author

Owner

Implementation Attempt — Tier 3: sonnet — Success

Fixed the suggestions() query extraction bug in on_input_submitted in src/cleveragents/tui/app.py.

Changes made:

Added import re to stdlib imports in app.py
Replaced text.replace("@", "").strip() with re.findall(r"@(\\S+)", text) to correctly extract only the last @token text (without the @ sign and without surrounding non-reference words) as the query passed to suggestions()
Created features/tdd_tui_suggestions_query_extraction_4741.feature with 4 TDD regression scenarios tagged @tdd_issue @tdd_issue_4741
Created features/steps/tdd_tui_suggestions_query_extraction_4741_steps.py with step definitions

Quality gate status: lint ✓, typecheck ✓, unit_tests ✓ (4 new scenarios + 24 existing pass)

PR: #10911

Automated by CleverAgents Bot
Supervisor: Implementation | Agent: task-implementor

**Implementation Attempt** — Tier 3: sonnet — Success Fixed the `suggestions()` query extraction bug in `on_input_submitted` in `src/cleveragents/tui/app.py`. **Changes made:** - Added `import re` to stdlib imports in `app.py` - Replaced `text.replace("@", "").strip()` with `re.findall(r"@(\\S+)", text)` to correctly extract only the last `@token` text (without the `@` sign and without surrounding non-reference words) as the query passed to `suggestions()` - Created `features/tdd_tui_suggestions_query_extraction_4741.feature` with 4 TDD regression scenarios tagged `@tdd_issue @tdd_issue_4741` - Created `features/steps/tdd_tui_suggestions_query_extraction_4741_steps.py` with step definitions **Quality gate status:** lint ✓, typecheck ✓, unit_tests ✓ (4 new scenarios + 24 existing pass) **PR:** https://git.cleverthis.com/cleveragents/cleveragents-core/pulls/10911 --- Automated by CleverAgents Bot Supervisor: Implementation | Agent: task-implementor

HAL9000 referenced this issue from a commit

2026-05-05 15:48:18 +00:00

fix(tui): add no-@token edge case scenario to TDD regression suite

HAL9000 referenced this issue from a commit

2026-05-07 15:26:38 +00:00

fix(tui): extract @token text correctly in on_input_submitted suggestions query

HAL9000 referenced this issue from a pull request that will close it,

2026-05-07 16:29:43 +00:00

fix(tui): extract @token text correctly in on_input_submitted suggestions query #11004