GCS - Mirage

The GCS resource mounts a Google Cloud Storage bucket at some prefix such as /gcs/. All operations involve network I/O to the remote object store. Uses aioboto3 against GCS’s S3-compatible XML API via HMAC keys, inheriting all S3 resource capabilities. For credential setup, see GCS Setup.

Config

import os

from mirage import MountMode, Workspace
from mirage.resource.gcs import GCSConfig, GCSResource

config = GCSConfig(
    bucket=os.environ["GCS_BUCKET"],
    access_key_id=os.environ["GCS_ACCESS_KEY_ID"],
    secret_access_key=os.environ["GCS_SECRET_ACCESS_KEY"],
    # Optional:
    # endpoint_url="https://storage.googleapis.com",
    # region="auto",
    # timeout=30,
    # proxy="http://proxy:8080",
)
resource = GCSResource(config)
ws = Workspace({"/gcs": resource}, mode=MountMode.READ)

GCSResource(config) takes a GCSConfig object with the bucket name and HMAC credentials. Both READ and WRITE modes are supported.

Filesystem Layout

The GCS resource maps object keys to virtual paths under the mount prefix, identical to the S3 resource. GCS “directories” are prefix-based — there are no real directory objects. For example, if bucket mirage-ai contains:

data/example.json
data/example.parquet
data/example.jsonl

Then mounting at /gcs/ exposes:

/gcs/
  data/
    example.json
    example.parquet
    example.jsonl

Path mapping: virtual /gcs/data/example.json maps to GCS key data/example.json.

Cache

The GCS resource uses IndexCacheStore with index_ttl = 600 (10 minutes), same as S3. Directory listings are cached for up to 600 seconds before being refreshed from GCS. This reduces API calls for repeated directory traversals.

Example

import asyncio
import os

from dotenv import load_dotenv

from mirage import MountMode, Workspace
from mirage.resource.gcs import GCSConfig, GCSResource

load_dotenv(".env.development")

config = GCSConfig(
    bucket=os.environ["GCS_BUCKET"],
    access_key_id=os.environ["GCS_ACCESS_KEY_ID"],
    secret_access_key=os.environ["GCS_SECRET_ACCESS_KEY"],
)

resource = GCSResource(config)


async def main() -> None:
    ws = Workspace({"/gcs/": resource}, mode=MountMode.READ)

    r = await ws.execute("ls /gcs/")
    print(await r.stdout_str())

    r = await ws.execute("cat /gcs/data/example.json | head -n 10")
    print(await r.stdout_str())

    r = await ws.execute("tree /gcs/")
    print(await r.stdout_str())

    r = await ws.execute("find /gcs/ -name '*.parquet'")
    print(await r.stdout_str())

    r = await ws.execute("stat /gcs/data/example.json")
    print(await r.stdout_str())


if __name__ == "__main__":
    asyncio.run(main())

Shell Commands

The GCS resource supports the full set of shell commands since it operates on real file content (text, binary, JSON, CSV, etc.). Large files benefit from range reads to avoid downloading entire objects.

Read Commands

Command	Notes
`cat`	Read file content
`head` / `tail`	First/last N lines
`grep` / `rg`	Pattern search (file or directory level)
`jq`	Query JSON fields
`wc`	Line/word/byte counts
`stat`	File metadata (name, size, type, modified)
`find`	Recursive search with `-name`, `-maxdepth`
`tree`	Directory tree view
`nl`	Number lines
`du`	Disk usage summary
`file`	Detect file type
`strings`	Extract printable strings from binary
`xxd`	Hex dump
`md5`	MD5 checksum
`sha256sum`	SHA-256 checksum

Text Processing

Command	Notes
`awk`	Pattern scanning and processing
`sed`	Stream editor
`tr`	Translate or delete characters
`sort`	Sort lines
`uniq`	Remove duplicate lines
`cut`	Extract fields/columns
`join`	Join lines on a common field
`paste`	Merge lines side by side
`column`	Columnate output
`fold`	Wrap lines to a specified width
`expand`	Convert tabs to spaces
`unexpand`	Convert spaces to tabs
`fmt`	Simple text formatter
`rev`	Reverse lines
`tac`	Concatenate and print in reverse
`look`	Display lines beginning with a given string
`shuf`	Shuffle lines
`tsort`	Topological sort
`comm`	Compare two sorted files
`cmp`	Compare two files byte by byte
`diff`	Compare files line by line
`patch`	Apply a diff patch
`iconv`	Character encoding conversion

File Operations

Command	Notes
`cp`	Copy files
`mv`	Move/rename files
`rm`	Remove files
`mkdir`	Create directories
`touch`	Create empty file or update timestamp
`ln`	Create symbolic links
`tee`	Write stdin to file and stdout
`mktemp`	Create temporary file
`split`	Split file into pieces
`csplit`	Split file by context

Path Utilities

Command	Notes
`basename`	Strip directory from path
`dirname`	Strip filename from path
`realpath`	Resolve path
`readlink`	Print symbolic link target
`ls`	List directory contents

Compression

Command	Notes
`gzip`	Compress files
`gunzip`	Decompress gzip files
`zip`	Create zip archives
`unzip`	Extract zip archives
`tar`	Archive files
`zcat`	Cat compressed files
`zgrep`	Grep compressed files

Encoding

Command	Notes
`base64`	Base64 encode/decode

Data Format Support

Commands with format-specific variants for structured data files:

Format	Extension	Variants
Parquet	`.parquet`	cat, head, tail, wc, stat, cut, grep, ls, file
Feather	`.feather`	cat, head, tail, wc, stat, cut, grep, ls, file
ORC	`.orc`	cat, head, tail, wc, stat, cut, grep, ls, file
HDF5	`.hdf5`	cat, head, tail, wc, stat, cut, grep, ls, file

These variants auto-detect the format by extension and convert to tabular text (CSV) for processing.

Use Cases

AI agents accessing GCS data: Mount GCS buckets for agents to read and process datasets
Data pipelines: Read and write GCS objects with shell-like commands
FUSE mounting: Expose GCS buckets through a virtual FUSE mount for external tools

​Config

​Filesystem Layout

​Cache

​Example

​Shell Commands

​Read Commands

​Text Processing

​File Operations

​Path Utilities

​Compression

​Encoding

​Data Format Support

​Use Cases