celeste-cliBLOCK
Agentic terminal AI assistant with a Bubble Tea TUI, code-graph review, semantic search, multi-provider LLMs, and MCP tools.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Character artwork by いかわさ (ikawasa23)
A premium, corruption-aesthetic command-line interface for CelesteAI
[](https://golang.org/) [](LICENSE) [](https://github.com/charmbracelet/bubbletea)
Built with [Charm's Bubble Tea](https://github.com/charmbracelet/bubbletea) for flicker-free, modern terminal experiences
✨ What is Celeste CLI?
Celeste CLI is a full standalone agentic development tool with her own persona, featuring:
- 🎨 Premium TUI - Flicker-free rendering with corrupted-theme aesthetics
- 🔮 48 Built-in Tools - File I/O, shell, web search, code graph, code review, collections search, git, crypto, subagent orchestration, and more
- 📖 `.grimoire` Project Context - Persona-themed project config files with auto-discovery;
AGENTS.md/CLAUDE.mdare read too - 🧠 Code Graph + Semantic Search - MinHash + BM25 fused ranking with LSH band table for sub-linear queries, structural rerank; tree-sitter parsing (TypeScript, PHP, Python, Rust, Java, C/C++ and Ruby) for accurate call-graph edges in release binaries (a
CGO_ENABLED=0source build falls back to regex parsers); embedded celeste-stopwords v1.0.0 noise filter - 🔍 Graph-Based Code Review - Structural analysis detecting stubs, lazy redirects, placeholders, error swallowing, and hardcoded values
- 🔌 Direct Codegraph MCP Tools -
celeste_index,celeste_code_search,celeste_code_review,celeste_code_graph,celeste_code_symbolsserved verbatim from the cached graph (no chat-LLM round-trip, nomax_tokens
8ee4b889ba98OBSERVED · 2026-10-11Exposed tools (2)
2 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
test | read | a test |
weather | read | get weather |
Trust audit
BLOCKgrade F · trust 52/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (7 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- found
Findings (25)
".ssh/id_", ".ssh/authorized_keys",
celeste_full.enc
celeste_lite.enc
celeste_off.enc
celeste_spine.enc
QExfiltration = "exfiltration"
var exfilPatterns = []struct {for _, p := range exfilPatterns {"DATABASE_URL=postgres://admin:[email protected]:5432/app": "s3cretpw",
const key = "ghp_abcdefghijklmnopqrstuvwxyz0123"
key := "-----BEGIN RSA PRIVATE KEY-----\n" + strings.Repeat("MIIEowIBAAKCAQEA", 200) + "\n-----END RSA PRIVATE KEY-----""-----BEGIN RSA PRIVATE KEY-----\nMIIEowIBAAKCAQEAtruncated": "MIIEowIBAAKCAQEAtruncated",
streamable-http
.gitleaks.toml
.golangci.yml
.grimoire
.release-please-manifest.json
content, isErr, err := exec()
echo "allow-loopback-pinentry" > ~/.gnupg/gpg-agent.conf
assert.NotEmpty(t, checkDangerousCommand("cat ~/.ssh/id_rsa"))assert.NotEmpty(t, checkDangerousCommand("cat ~/.aws/credentials"))assert.NotEmpty(t, checkDangerousCommand("cat ~/.kube/config"))_, err = store.Load("../../etc/passwd"){"sub/../../run1", true},{"../../etc/passwd", true},Gates applied: no_behavioural_pass.
8ee4b889ba98full audit observations/trust-audit/mcp-server/whykusanagi__celeste-cli.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 8ee4b889ba98 | BLOCK | F | 52 | first audit |
Questions
What is the celeste-cli MCP server?
Agentic terminal AI assistant with a Bubble Tea TUI, code-graph review, semantic search, multi-provider LLMs, and MCP tools.
What tools does celeste-cli expose?
2 in total: 2 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is celeste-cli safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (52/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.
What credentials does celeste-cli need?
No credential environment variables were found in its source, so it appears to need none.
How does celeste-cli run?
It speaks streamable-http, so it runs as a service you connect to over the network.
How current is this page?
The grade is for one exact copy of the source (8ee4b889ba98), read on 2026-10-11. The repository is watched and re-audited when it changes.