Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/rsafargalin/sextant/code-navigationnpx skills add RSafargalin/sextant --skill code-navigationgit clone --depth 1 https://github.com/RSafargalin/sextantWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00074 | $0.00664 |
| Opus 5 | $0.00037 | $0.00332 |
| Sonnet 5 | $0.00015 | $0.00133 |
| Haiku 4.5 | $0.00007 | $0.00066 |
Grade A, and why
code-navigation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Navigating a compiled codebase
The project's build already wrote an index store — the data behind "jump to definition" in an
editor. The sextant MCP tools read it. An exact answer is one call away, so a grep-and-read
loop over the same question is both slower and wrong more often: a text search finds a name in a
comment, misses a call spelled through a protocol, and cannot see across the Swift ↔ Objective-C
boundary.
Ask the index, not the filesystem
| The question | What to do |
|---|---|
Where is X defined? Who uses it? What does it call? |
One context call — not a Grep followed by three Reads |
What breaks if I change X? |
blast_radius |
| Who calls this function? | find_callers — it accounts for protocol dispatch, which grep cannot |
| What implements this protocol / subclasses this type? | list_implementations |
| What does this package or type expose? | api — an order of magnitude cheaper than reading the sources |
Where is this code shape used (try? $X.save(), [$X reloadData])? |
structural_search — an AST pattern, not a regex |
| What changed between two revisions? | changed — declaration-level, not a line diff |
| I need the implementation, not the signature | body |
Reach for Grep or a file read when the target is not a symbol: a string literal, a comment, a build setting, a resource, a file whose name you already know.
Read the index line of every answer
Every answer carries a line naming the store it was read from and whether it is fresh — the last line of the result, in brackets. Treat it as part of the answer:
- stale — the store predates the current sources; a symbol added since the last build is absent from it. Say so rather than reporting "no references".
- no index — the semantic tools return a hint instead of an empty result. Building one runs
the project's own build (
sextant index), and that is the user's decision to make, not a step to take on their behalf. Ask. repo_map,structural_searchandlintneed no index for Swift; for Objective-C, C and C++ they need the compile flags thatsextant indexrecords.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 47 lines · 74 tokens per session scan A bb3319970f8c
code-navigation is a skill published in the GitHub repository RSafargalin/sextant (2 stars, last pushed 15d ago), licensed Apache-2.0. It adds 74 tokens to every session and 664 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
release-codexbar
CodexBar release: versioning, notarization, appcast, Homebrew, post-release bump.
codexbar
CodexBar read. Provider usage, limits, credits, config health. JSON. No writes.
release-peekaboo
Peekaboo release: notarization, npm/GitHub release, appcast, verify, closeout.
review-knowledge
Audit the knowledge/ base and the .claude/ skills, agents and workflows for staleness, self-contradiction, and drift from their own stated rules — four independent adversarial auditors (two lenses over two trees) verify every load-bearing claim against the current tree, cross-examine each other within their tree, and…
implement-plan
Implement the current plan test-first, the Canon TDD way, driving to a single finishing condition — an empty test list. Derives a Canon TDD test list from the plan, shows it before any code, writes one failing test at a time, and keeps the list visible as it evolves. Use when the user asks to implement, build, or…
cut-release
Cut a new TMDb release — work out the next SemVer version from the evidence, do the pre-tag housekeeping a tag would otherwise freeze in place, draft release notes, then tag and publish the GitHub release. Presents the version (with its reasoning) and the full notes for approval and STOPS; nothing is tagged or…