Q doesn't seem to not load context files

### Checks

- [x] I have searched [github.com/aws/amazon-q-developer-cli/issues](https://github.com/aws/amazon-q-developer-cli/issues?q=) and there are no duplicates of my issue
- [x] I have run `q doctor` in the affected terminal session
- [x] I have run `q restart` and replicated the issue again

### Operating system

macOS 15.4.1 (24E263)

### Expected behaviour

```
> /context show

🌍 global:
.amazonq/rules/**/*.md
README.md (1 match)
AmazonQ.md
CLAUDE.md (1 match)
/Users/david/.dotfiles/llm/q_rules.md (1 match)

👤 profile (default):

4 matched files in use:
🌍 /Users/david/src/xxx/README.md (~840 tkns)
🌍 /Users/david/src/xxx/CLAUDE.md (~7640 tkns)
🌍 /Users/david/.dotfiles/llm/q_rules.md (~1090 tkns)
```

From this I am taking that Q properly loaded: CLAUDE.md, my `q_rules.md` and `README.md`

However when asked Q about anything written in those files, it got nothing:

```
> tell me what you know about this project without reading any files

Based on the system context information I have, I can tell you that:

This project appears to be called "xxx" and is located in the directory "/Users/david/src/xxx" on a
macOS system.

The project seems to be integrated with Linear (a project management tool), as there are several Linear-
related functions available that suggest this project has capabilities for:
• Managing projects and milestones
• Creating and tracking issues
• Managing comments on issues
• Working with teams
• Searching and filtering issues

Without examining any files, I can't provide specific details about the implementation, architecture, or
purpose of the project beyond these observations. Would you like me to explore the project files to provide
more detailed information?
```

Then in my `q_rules.md` I have some extra capabilities, a compacted version looks like this:

```
❯ cat ~/.dotfiles/llm/q_rules.md
===

Extra capabilities

- You have `trafilatura` as CLI tool available, to fetch the text content of websites. Use this to research specific URLs when provided. Example: `trafilatura -u https://developers.cloudflare.com/cloudflare-one/connections/connect-networks/private-net/cloudflared/`
```

```
> do you see any mention of trafilatura in your context?

No, I don't see any mention of "trafilatura" in the context provided. The context entry doesn't contain any
specific information about trafilatura, which is a Python package used for web scraping that focuses on
extracting text content from web pages.

If you're interested in using trafilatura for web scraping or text extraction, I'd be happy to help you with
that. Would you like information about how to use trafilatura or assistance with a specific task involving
it?
```

So it seems like none of the context files are actually loaded into Q context

### Actual behaviour

- Q should have all context matches as part of it's context

### Steps to reproduce

_No response_

### Environment

```yaml
[q-details]
version = "1.9.1"
hash = "5d2c2807cab9bc6c9dc5fcd05f92c7a979e8a4f7"
date = "2025-04-29T06:22:42.917989Z (1d ago)"
variant = "full"

[system-info]
os = "macOS 15.4.1 (24E263)"
chip = "Apple M1 Max"
total-cores = 10
memory = "64.00 GB"

[environment]
cwd = "/Users/USER/src/microfn"
cli-path = "/Users/USER/src/microfn"
os = "Mac"
shell-path = "/opt/homebrew/Cellar/fish/3.7.1/bin/fish"
shell-version = "3.7.1"
terminal = "Ghostty"
install-method = "brew"
SHELL = "/opt/homebrew/bin/fish"
TERM = "xterm-ghostty"
__CFBundleIdentifier = "com.mitchellh.ghostty"

[chat-settings]
interactive=true

[chat-trusted_tools]

[chat-failed_request_ids]
none

[chat-context]
current_profile=default
profiles=
default
aiquickshot

global_context=
.amazonq/rules/**/*.md
README.md
AmazonQ.md
CLAUDE.md
/Users/david/.dotfiles/llm/q_rules.md

files=none
```

Riproduci il report con `q restart`, `q doctor` e `/context show`, quindi confronta le corrispondenze elencate con l'output dell'ambiente `[chat-context]` che mostra `files=none`. Traccia il motivo per cui i file corrispondenti vengono visualizzati ma non inclusi nel contesto della chat; il lavoro è completato quando i file sono sia elencati sia disponibili per le domande nella sessione.

Scritto dal modello di indicizzazione a partire dal testo della issue.