1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
// Copyright 2025 Muvon Un Limited
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
// Hint accumulator — collects tool-misuse hints during a tool execution round.
//
// Tools call push_hint() instead of appending ⚠️ text to their result content.
// After all parallel tool results are collected, the session layer calls drain_hints()
// and injects a single user-role message so the AI sees the guidance without
// polluting individual tool result strings.
//
// Thread-local storage is correct here: each tokio::spawn task gets its own
// OS thread (or is pinned to one), so hints from concurrent tool calls accumulate
// independently and are drained by the coordinating task after join_all().
// The drain happens in the same task that spawned the tool tasks, which runs on
// the coordinator thread — so we use a Mutex-protected global instead to safely
// collect across spawn boundaries.
use Mutex;
static HINTS: = new;
/// Push a hint into the accumulator. Deduplication is applied at drain time.
/// Only call this when the recommended tool is actually enabled (check before calling).
/// Drain all accumulated hints, returning deduplicated list in insertion order.
/// Clears the accumulator — ready for the next tool execution round.
/// Returns true if there are any pending hints (without draining).