atomwrite 0.1.38

# atomwrite

> Atomic file operations for LLM agents: one CLI with 47 subcommands, NDJSON on stdout and BLAKE3 checksums on every write

atomwrite is a Rust CLI for reading, writing, editing, searching and transforming files. Every write uses tempfile, fsync, rename and a directory fsync, so a crash never leaves a half-written file. Every response is NDJSON, one JSON object per line. Current version is 0.1.38 (MSRV 1.90, pin `^0.1.38`), with no telemetry and no product environment variables: configuration comes from flags, the workspace `.atomwrite.toml` and the global XDG `config.toml`.

Integration is by subprocess. Run the CLI, read the exit code first, then parse stdout. Set `WS` once to the absolute workspace root and pass `--workspace "$WS"` to every file command, because the jail is anchored on that value. The full reference, with every flag and every example expanded, is in `llms-full.txt`.

## Documentation
- [README](https://github.com/danilo-aguiar-br/atomwrite/blob/main/README.md): install, all 47 subcommands, every local flag per subcommand and examples
- [HOW_TO_USE](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/HOW_TO_USE.md): quickstart and one block per subcommand
- [AGENTS](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/AGENTS.md): agent discovery file with the output contract and the token budget
- [COOKBOOK](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/COOKBOOK.md): recipes for common tasks
- [INTEGRATIONS](https://github.com/danilo-aguiar-br/atomwrite/blob/main/INTEGRATIONS.md): integration patterns for agents and scripts
- [ARCHITECTURE](https://github.com/danilo-aguiar-br/atomwrite/blob/main/ARCHITECTURE.md): how the write path, the jail and the WAL work
- [INSTALL](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/INSTALL.md): platform-specific install instructions and Cargo features
- [llms-full.txt](https://github.com/danilo-aguiar-br/atomwrite/blob/main/llms-full.txt): the expanded version of this file

## Output Contract
- Exit code first: read the process exit code before you parse stdout
- stdout is NDJSON only, one JSON object per line, and stderr carries logs only, so never parse stderr for data
- `--json` is accepted and ignored, because the output is always NDJSON
- Read the `type` of the last line: query verbs end with a summary event and write verbs end with their own result event
- `read`, `write` and `hash` expose `checksum`; `edit` exposes `checksum_before` and `checksum_after`
- A failed run ends with exactly one terminal event: `{"ok":false,"code":"...","exit":N,"path":"...","suggestion":"..."}`, and `message`, `error_class`, `retryable` and `suggestion_lang` appear only under `-v`
- `error_class` is `permanent`, `transient`, `conflict` or `precondition_failed`; retry only when `retryable` is true, and both fields stay visible only under `-v`
- `MATCH_FAILED` and `MATCH_AMBIGUOUS` (exit 65) add `best_candidate` and `match_count`; read `best_candidate` before you change `--old`
- A cancelled run ends with `{"type":"cancelled","reason":"...","signal":"...","deadline_exceeded":false}`, and `deadline_exceeded` separates the `--timeout-secs` deadline (exit 124) from a signal (exit 130 or 143)
- `read` on a picture answers one `data:` URI line on stdout, and `--max-tokens`, `--crop` and `--high-res` choose the cost before the encode
- `read` on a PDF answers the text of each page, and a page with no text is named in `needs_vision`, because a scanned page is not an empty document
- `read doc.pdf --page N` renders that page as a `data:` URI, `--dpi N` sets the resolution, and `--ocr` returns the recognized text in `content`
- `--select`, `--filter` and `--limit` never hide an `ok:false` envelope
- Zero matches in `search`, `replace`, `transform`, `scope` and `semantic-search` exit 1 with `NO_MATCHES`, which is not a crash, and `search` exits 89 with `ALL_SKIPPED_BINARY` when every candidate file was binary
- Every verb that walks a tree lists unreadable files in `files_failed`, `partial` and `errors[]`, so read them before you trust a result
- Verbs that write need an explicit path (exit 2 without it) and report `target_resolved` and `target_source` on the terminal event, where `target_resolved` is always an array even for one path, and `target_source` is `argv`, `workspace` for an idempotent read verb that fell back to the root, or `stdin` for the `-` operand
- Every terminal event also carries `workspace_source`, which is `argv` when the root came from `--workspace` and `cwd` when it came from the process working directory
- The JSON Schemas of this contract are in [docs/schemas/](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/schemas/README.md), where 54 files describe the 55 names `atomwrite --print-schema` lists

## Exit Codes
- 0: success
- 1: `NO_MATCHES` (zero results), or `DOCTOR_CHECKS_FAILED` (`doctor --strict` with a failed check)
- 2: invalid command line (`ARGUMENT_PARSE_ERROR`), or flags that cannot run together (`UNSUPPORTED_COMBINATION`)
- 4: file not found (`FILE_NOT_FOUND`), or a glob operand that matched nothing (`GLOB_NO_MATCHES`)
- 13: `PERMISSION_DENIED`
- 28: disk full (transient, retryable)
- 30: quota exceeded (transient, retryable)
- 65: `INVALID_INPUT`, `MATCH_FAILED`, `MATCH_AMBIGUOUS`, `BINARY_FILE`, `FILE_TOO_LARGE`, a missing `--ack-overwrite`, a file above `[fuzzy].max_content_bytes` under fuzzy, an invalid `config` key or value
- 69: `GRAMMAR_UNAVAILABLE`, an AST verb with no cached parser and no network
- 73: cross-device rename (conflict, retryable)
- 74: I/O error (transient, retryable)
- 78: `CONFIG_INVALID` (unknown key or invalid file), or a verb whose build feature is missing, such as `watch`
- 79: `SYMLINK_BLOCKED`, a symlink the jail refuses; `--follow-symlinks` resolves it and checks the target against the workspace
- 80: batch transaction rollback failed
- 81: checksum verify failed
- 82: `STATE_DRIFT`, the file changed since it was read (conflict, retryable)
- 83: `LOCK_TIMEOUT`, the automatic write lock waited 5 s (conflict, retryable)
- 84: semantic-merge conflict
- 85: FIFO detected
- 86: device file detected
- 87: `edit-loop --partial` wrote the matching pairs and some pairs missed
- 88: syntax error detected by `--syntax-check`
- 91: EXDEV fallback disabled
- 92: copy-back BLAKE3 check failed (conflict, retryable)
- 93: orphan journal detected
- 124: global `--timeout-secs` deadline exceeded (default 120, `0` disables)
- 126: workspace jail violated
- 128: file immutable
- 129: SIGHUP
- 130: SIGINT or cancelled
- 141: broken pipe, the reader closed the pipe and the output is incomplete (cooperative handler, cleanup runs first)
- 143: SIGTERM, a clean cancel with no tempfile left
- 255: internal error

## Global Flags
- Workspace and config: `--workspace <DIR>`, `--config <PATH>`, `--locale <en|pt-BR>`, `--no-auto-heal`, `--ready-file <PATH>` (test harness only)
- Limits: `--timeout-secs <N>` (default 120, alias `--timeout`), `-j`/`--threads <N>` (alias `--max-concurrency`, default automatic up to 16, capped by RAM), `--max-filesize <BYTES>` (default 1 GiB, larger files are rejected; `search` has its own 10 MiB skipping limit)
- Walk: `--no-gitignore`, `--hidden`, `--follow-symlinks`
- Output: `-v`/`-vv`/`-vvv`, `-q`/`-qq`, `--color auto|always|never`, `--no-color`, `--no-progress`
- Schema: `--json-schema`, `--print-schema [NAME]` (55 names, put it before the command, an unknown name exits 65)
- Agent mode: `--non-interactive` (never prompt), `--confirm-side-effect` (only where the command contract asks), `--allow-truncate`
- Shape: `--select`/`--fields KEYS`, `--filter key=value` (repeatable AND), `--limit N` as a work budget: the target list is cut before the side effect, so `--limit 1` touches exactly one target (long-only, `-n` belongs to `replace`)
- Stream shape: `--sort-by path-line|line-path|KEY`, `--dedupe KEYS`, `--aggregate KEY[:OP[:FIELD]]`, `--join KEY:SUFFIX`, `--diff-against PATH`, `--truncate-content KEY:LEN`, `--redact`, `--normalize-encoding`
- `--max-output-bytes N` cuts every string value, `checksum` included, and needs `--allow-truncate` (each is rejected without the other, exit 2)

## Commands
- `read` (Core I/O): read a file with checksum, size and a bounded slice; `--number`, `--symbol`, `--line N --enclosing`, repeatable `--lines` and `--all` pick the slice, and a cut answers `truncated: true` with the `next` hint. Example: `atomwrite --workspace "$WS" read --head 2 src/a.rs`
- `write` (Core I/O): create or overwrite a file atomically from stdin. Example: `printf 'x\n' | atomwrite --workspace "$WS" write notes/new.txt`
- `edit` (Core I/O): edit by exact or fuzzy text, line number or marker. Example: `atomwrite --workspace "$WS" edit src/b.rs --old beta --new gamma`
- `delete` (Core I/O): delete files or trees with backup, or print the plan with `--plan` (a directory needs `-r`, and a tree delete publishes `backups_kept`, `backup_bytes` and `dirs_left` when a backup or a leftover exists). Example: `atomwrite --workspace "$WS" delete --plan tmp.txt`
- `copy` (Core I/O): copy a file or tree with checksum verification. Example: `atomwrite --workspace "$WS" copy src/a.rs src/a_copy.rs`
- `move` (Core I/O): move or rename a file atomically. Example: `atomwrite --workspace "$WS" move src/tmp.rs src/tmp2.rs`
- `apply` (Core I/O): apply a patch (unified diff, SEARCH/REPLACE, full file) from stdin; the target is optional, so one call patches several files, and `--show` returns the changed lines. Example: `printf 'new body\n' | atomwrite --workspace "$WS" apply --format full notes.txt`
- `search` (Search and replace): search file contents in parallel with the ripgrep engine; `--format text` prints one `path:line:col:text` per match. Example: `atomwrite --workspace "$WS" search alpha src/`
- `replace` (Search and replace): replace text across files with atomic writes and fuzzy matching. Example: `atomwrite --workspace "$WS" replace --dry-run alpha omega src/`
- `transform` (Search and replace): rewrite code by AST pattern with ast-grep. Example: `atomwrite --workspace "$WS" transform --dry-run -p '$E.unwrap()' -r '$E?' -l rust src/`
- `hash` (Inspection): print the BLAKE3 checksum of files. Example: `atomwrite --workspace "$WS" hash src/a.rs`
- `count` (Inspection): count lines and files by extension or by size. Example: `atomwrite --workspace "$WS" count --by-extension src/`
- `diff` (Inspection): compare two files (unified, stat or changes); each change carries a 1-based `line` plus `old_line` and `new_line`. Example: `atomwrite --workspace "$WS" diff src/a.rs src/b.rs --unified`
- `list` (Inspection): list a directory tree with metadata; `--format text` prints one path per line. Example: `atomwrite --workspace "$WS" list --long --depth 1 src/`
- `extract` (Inspection): pick fields from NDJSON read on stdin, with the local `--field` and never `--select`. Example: `printf '{"a":1}\n' | atomwrite extract --field a`
- `extract-artifact` (Inspection): bounded signal from PDF, OOXML and image headers, never raw bytes. Example: `atomwrite --workspace "$WS" extract-artifact img.png`
- `scope` (Inspection): grammatical scoping, such as deleting all comments. Example: `atomwrite --workspace "$WS" scope src/ --language rust --query comments --delete --dry-run`
- `regex` (Inspection): generate a regex from example strings. Example: `atomwrite regex abc-12 xyz-99 id-7 --digits`
- `calc` (Inspection): math and unit conversion. Example: `atomwrite calc "2 hours to seconds"`
- `completions` (Inspection): print or install shell completions for bash, zsh, fish, elvish and powershell. Example: `atomwrite completions bash`
- `backup` (Backup and recovery): create timestamped backups with BLAKE3 checksums. Example: `atomwrite --workspace "$WS" backup src/a.rs --retention 3`
- `rollback` (Backup and recovery): restore a file from a previous backup. Example: `atomwrite --workspace "$WS" rollback src/a.rs --latest --verify`
- `batch` (Backup and recovery): run NDJSON operations from stdin, optionally as one transaction. Example: `printf '{"op":"write","target":"batch.txt","content":"x"}\n' | atomwrite --workspace "$WS" batch --transaction`
- `edit-loop` (Batch helpers): apply N old/new pairs in one write, all or nothing (`--partial` exits 87). Example: `printf '[{"old":"gamma","new":"delta"}]\n' | atomwrite --workspace "$WS" edit-loop src/b.rs`
- `prune-backups` (Batch helpers): remove old `.bak` files by age or count. Example: `atomwrite --workspace "$WS" prune-backups -r --max-age-secs 86400 --dry-run .`
- `wal-stats` (WAL diagnostics): show the state of the WAL journals, read only. Example: `atomwrite --workspace "$WS" wal-stats`
- `wal-heal` (WAL diagnostics): remove orphan terminal WAL journals older than a threshold, and also remove a `Started` journal whose owning PID is gone, reported as `dead_owner_removed`. Example: `atomwrite --workspace "$WS" wal-heal --threshold-secs 3600 .`
- `verify` (Verification): check a file against an expected BLAKE3 hash, exit 0 on match and 81 on mismatch. Example: `atomwrite --workspace "$WS" verify src/a.rs "$HASH_A"`
- `set` (Structured editors): write a value at a dotted path in a TOML or JSON file. Example: `atomwrite --workspace "$WS" set config.toml package.name demo2`
- `get` (Structured editors): read a value at a dotted path in a TOML or JSON file. Example: `atomwrite --workspace "$WS" get config.toml package.name`
- `del` (Structured editors): remove a key from a TOML or JSON file. Example: `atomwrite --workspace "$WS" del config.toml features.x`
- `case` (Structured editors): rename identifiers across files and directories between case styles, and a directory is expanded through the walker. Example: `atomwrite --workspace "$WS" case --subvert user_id UserId --to pascal --dry-run src/c.rs`
- `query` (AST tools): walk the tree-sitter AST and emit nodes. Example: `atomwrite --workspace "$WS" query --kinds src/a.rs`
- `outline` (AST tools): emit the high-level structure of one source file; `--format tsv` prints `line<TAB>kind<TAB>name` with the path once. Example: `atomwrite --workspace "$WS" outline src/a.rs`
- `recipe` (Agent surface): list and run built-in multi-step recipes. Example: `atomwrite --workspace "$WS" recipe run --name search-replace-verify --pattern alpha --replacement omega --path src --dry-run`
- `sparse` (Agent surface): budgeted list, read and outline for large trees. `--all` adds the hidden entries to the visit. Example: `atomwrite --workspace "$WS" sparse list src/ --max-files 10 --max-bytes 1048576 --all`
- `semantic-merge` (Agent surface): line-based three-way merge with conflict markers, not AST. Example: `atomwrite --workspace "$WS" semantic-merge --base base.txt --ours ours.txt --theirs theirs.txt --output merged.txt --fail-on-conflict`
- `agent-surface` (Agent surface): tool manifest for agents, the single source of truth for the subcommands. Example: `atomwrite agent-surface --format json`
- `watch` (Agent surface): stream filesystem events with debounce, ends with `watch_summary`, needs a build with `--features watch` (the default binary exits 78). Example: `atomwrite --workspace "$WS" watch src/ --max-events 1 --idle-exit-ms 500`
- `codemod` (Agent surface): apply a YAML list of AST rewrite rules across a tree. Example: `atomwrite --workspace "$WS" codemod --rules rules.yaml --dry-run src/`
- `semantic-search` (Agent surface): offline Jaccard ranking of files for a query. Example: `atomwrite --workspace "$WS" semantic-search "unwrap value" src/ --k 5 --min-score 0.01`
- `stat` (Agent surface): metadata only, alias of `read --stat`. Example: `atomwrite --workspace "$WS" stat src/a.rs`
- `doctor` (Agent surface): host readiness diagnostics, `--strict` exits 1 on a failed check. Example: `atomwrite --workspace "$WS" doctor`
- `locale` (Agent surface): show, set or clear the UI locale. Example: `atomwrite locale`
- `commands` (Agent surface): print the full command tree as JSON; `--brief` prints one line per subcommand. Example: `atomwrite commands --include-globals`
- `scratch` (Agent surface): create the staging directory inside the workspace jail, never the OS `/tmp`. Example: `atomwrite --workspace "$WS" scratch init .atomwrite/scratch`
- `config` (Agent surface): read and write persistent tuning keys with `path|list|get|set|unset|doctor`. Example: `atomwrite --workspace "$WS" config get fuzzy.threshold`

## Configuration
- 43 keys, listed by `atomwrite config list`: `defaults.line_ending`, `defaults.backup`, `defaults.retention`, `defaults.max_filesize`, `delete.backup`, `fuzzy.mode`, `fuzzy.threshold`, `fuzzy.threshold_aggressive`, `fuzzy.threshold_context`, `fuzzy.threshold_jw`, `fuzzy.soft_floor_delta`, `fuzzy.max_content_bytes`, `apply.max_ellipsis_lines`, `write.confirm_large_bytes`, `write.shrink_block_percent`, `write.auto_rotate_max_age_secs`, `write.content_risk_patterns`, `watch.idle_exit_ms`, `watch.debounce_ms`, `log.file`, `log.rotation`, `storage.home`, `locale`, `concurrency.max_threads`, `concurrency.ram_fraction_percent`, `search.context`, `search.smart_case`, `search.max_filesize`, `search.max_columns`, `search.max_inflight_bytes`, `search.format`, `read.max_output_bytes`, `read.number`, `output.profile`, `summary.errors_max_entries`, `lock.acquire_timeout_ms`, `retry.base_ms`, `retry.max_ms`, `retry.max_elapsed_ms`, `extract.max_sample_rows`, `extract.max_text_chars`, `extract.max_sheets`, `extract.max_slides`
- `read.max_output_bytes` (24000) bounds `read` and `--all` removes it, `search.format` (`ndjson` or `text`) picks the default search shape, `output.profile` (`full` or `lean`) trims a mutation record, and `apply.max_ellipsis_lines` (200) bounds a `...` span
- Precedence: flags, then the workspace `.atomwrite.toml`, then the global XDG `config.toml` (the two files are merged key by key), then built-in defaults
- `config set KEY VALUE --scope global|workspace` and `config unset KEY --scope ...` need `--scope`, while `config list` and `config get KEY` do not accept it and show `value`, `origin` and `default`
- `storage.home` and `locale` are global only, while `log.file` and `log.rotation` are accepted in both scopes
- An unknown key or an invalid file exits 78, and an invalid `config set` key or value exits 65 without writing

## Upgrading To 0.1.38
- The default `read` stops at `read.max_output_bytes` (24,000 bytes), cuts on a complete line and prints the exact next `--lines` on stderr, while `--all` writes the whole file
- `--number`, `--symbol NAME`, `--line N --enclosing`, repeatable `--lines` with `--range` as an alias and `--no-comments` are the cheap read shapes, and `--no-comments` without `--number` exits 65
- `search --format text`, `list --format text` and `outline --format tsv` are the compact shapes, and `search.format` makes the text search the default
- `apply` patches several files in one call when the patch names them, matches whole lines, refuses a repeated SEARCH with exit 65 `MATCH_AMBIGUOUS` unless `--all` is given, and runs the syntax check before and after the swap
- `apply --format unified` accepts `@@ ... @@` and a bare `@@`, matches a hunk without numbers by context, and refuses more than one match by naming the line of each
- `target_resolved` is always an array, `error: true` left the failure envelope, and `message`, `error_class`, `retryable` and `suggestion_lang` appear only under `-v`
- `delete -r <dir>` owns the whole named tree, and `delete --backup` with a directory root is refused before the first unlink
- The lock is taken on the target inode with `std::fs::File::try_lock`, and the `.<target>.atomwrite.lock` sidecar is best effort
- `config` lists 43 keys, `output.profile = "lean"` trims a mutation record, and `atomwrite commands --brief` prints one line per subcommand
- Details are in [MIGRATION](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/MIGRATION.md)


## Upgrading To 0.1.38 (read expansion)
- `read` takes one or more paths: explicit paths in argv order, and an operand with `*`, `?` or `[` is a pattern the binary expands against one level of its own directory, in `LC_ALL=C` byte order, skipping hidden names
- `--format raw` concatenates every target byte for byte like `cat`, with no header, no separator and no added newline, and copies in 64 KiB blocks so memory does not grow with the batch
- A failing target prints one line on stderr and the run continues, and the final status is the semantic code of the first failure (4 missing, 13 permission, 126 outside the `--workspace`)
- A `read --format raw` error goes to stderr, never to stdout, because that channel promised bytes
- `--lines`, `--head`, `--tail`, `--context`, `--grep` and `--no-comments` work on several targets in `--format raw` and print one `==> path <==` header per file, while `--stat` and `--verify-checksum` need a single target and exit 65 with more than one
- `--format raw` stops at 24,000 bytes unless `--all` is passed, and prints the next `--lines` on stderr
- A pattern with `**` or with more than one separator exits 65, and the default NDJSON format is unchanged
- Details are in [MIGRATION](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/MIGRATION.md)


## Upgrading To 0.1.37 (draft of 2026-10-15)
- Risk assessment now runs in every verb that writes: `edit`, `edit-loop`, `replace`, `case`, `transform` and `codemod` publish `risk_assessment` when the payload matches a content-risk pattern
- A tree-sitter grammar this host cannot provide is exit 69 `GRAMMAR_UNAVAILABLE` with class `precondition_failed`, never 255 `INTERNAL_ERROR`
- `write.content_risk_patterns` is a list value in `config get` / `config set`, and the config table holds 43 keys
- `log.file` and `log.rotation` accept the workspace scope, and the `[log]` section of `.atomwrite.toml` starts the log sink
- `batch --transaction` publishes one verdict: the `rollback` event carries `ok: false` with the failure's `code` and `exit`
- `--version` no longer appends `-dirty`, so an untracked file cannot change the string
- Details are in [MIGRATION](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/MIGRATION.md)


## Upgrading To 0.1.37 (agent contract)
- `watch` left the default build: install with `cargo install atomwrite --features watch`, otherwise it exits 78
- `edit-loop` is all-or-nothing: a pair with no match exits 65 and writes nothing, and `--partial` exits 87
- A failing summary is the only error line of the run, so read the summary instead of waiting for a second error line
- `apply` with a unified diff compares context and removed lines, so a mismatch exits 65 `MATCH_FAILED` and writes nothing
- `apply` takes several files in one call: omit the target and name each file with a bare path line above its first block, or send a unified diff whose header pairs name each file
- `semantic-merge` with a conflict exits 84, and `case`, `transform`, `scope` and `codemod` can exit 82 `STATE_DRIFT`
- An invalid regex or glob exits 65 instead of 255
- `--print-schema` must come before the command name
- `replace`, `scope`, `codemod`, `case`, `delete`, `backup`, `prune-backups`, `wal-heal`, `scratch init` and `recipe run` need an explicit path
- Details are in [MIGRATION](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/MIGRATION.md)

## Quality
- 131 test targets passed at the 0.1.37 release gate, and 55 stable JSON Schema names back the output contract, served by 54 files
- 77 architecture decision records live in [docs/decisions/](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/decisions/) (0019 to 0095)
- Tree-sitter parsers are downloaded on first use, honour the proxy settings of the host and are cached under the XDG cache directory
- The first run of each language therefore needs the network, `atomwrite doctor` publishes the check `grammar_cache` with how many of the 23 grammars the AST verbs can need are already cached, and a grammar that cannot be obtained is exit 69 `GRAMMAR_UNAVAILABLE`, never 255
- A panic writes a local report file in the temporary directory and nothing is ever transmitted

## Optional
- [CHANGELOG](https://github.com/danilo-aguiar-br/atomwrite/blob/main/CHANGELOG.md): release history
- [MIGRATION](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/MIGRATION.md): upgrade guide between versions
- [CROSS_PLATFORM](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/CROSS_PLATFORM.md): macOS, Linux and Windows quirks
- [TESTING](https://github.com/danilo-aguiar-br/atomwrite/blob/main/docs/TESTING.md): how the test suites are organized
- [PRIVACY](https://github.com/danilo-aguiar-br/atomwrite/blob/main/PRIVACY.md): what the CLI reads, writes and never sends
- [SECURITY](https://github.com/danilo-aguiar-br/atomwrite/blob/main/SECURITY.md): vulnerability reporting
- [CONTRIBUTING](https://github.com/danilo-aguiar-br/atomwrite/blob/main/CONTRIBUTING.md): contributor onboarding, local definition of done and release process
- [CODE_OF_CONDUCT](https://github.com/danilo-aguiar-br/atomwrite/blob/main/CODE_OF_CONDUCT.md): Contributor Covenant 2.1 community standards
- [Crates.io](https://crates.io/crates/atomwrite): package registry
- [Repository](https://github.com/danilo-aguiar-br/atomwrite): source code