Spaces:
Running on Zero
Running on Zero
File size: 13,489 Bytes
5f43c7d 2c43c61 5f43c7d 5048a89 5f43c7d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 | #!/usr/bin/env python3
"""Her · हेर — bulk session uploader (scan → scrub → upload, with your approval).
Brings your Claude Code sessions into the private Her Space so you get a full Projects
view. It NEVER touches your originals: it COPIES the sessions you pick into a local
staging folder, SCRUBS likely secrets from the copies, then UPLOADS them — pausing for
your approval at each of the three steps.
Pure standard library — no pip installs. Run:
python her_upload.py
python her_upload.py --space build-small-hackathon/her # override the Space
python her_upload.py --projects-dir ~/.claude/projects # override the source
Auth: uses your Hugging Face token (HF_TOKEN env, else ~/.cache/huggingface/token —
created by `hf auth login`). Required because the Space is private.
PRIVACY: the scrubber is best-effort (you review the redaction summary before upload),
and your uploads auto-delete from the Space after 24h (or when you click "clear my data"
/ close the tab). Nothing here ever modifies ~/.claude.
"""
from __future__ import annotations
import argparse
import glob
import json
import os
import re
import shutil
import sys
import uuid
import urllib.request
import urllib.error
from pathlib import Path
DEFAULT_SPACE = "build-small-hackathon/her"
# --------------------------------------------------------------------------- #
# small console helpers
# --------------------------------------------------------------------------- #
def c(txt, color="orange"):
codes = {"orange": "38;5;208", "red": "31", "green": "32", "cyan": "36", "dim": "2", "bold": "1"}
return f"\033[{codes.get(color,'0')}m{txt}\033[0m"
def hr():
print(c("─" * 64, "dim"))
def ask(prompt: str) -> str:
try:
return input(prompt).strip()
except (EOFError, KeyboardInterrupt):
print("\naborted.")
sys.exit(1)
def confirm(prompt: str) -> bool:
return ask(prompt + " [y/N] ").lower() in ("y", "yes")
# --------------------------------------------------------------------------- #
# auth + host
# --------------------------------------------------------------------------- #
def hf_token() -> str:
tok = os.environ.get("HF_TOKEN") or os.environ.get("HUGGING_FACE_HUB_TOKEN")
if tok:
return tok.strip()
for p in (Path.home() / ".cache/huggingface/token", Path.home() / ".huggingface/token"):
try:
t = p.read_text(encoding="utf-8").strip()
if t:
return t
except OSError:
pass
print(c("No Hugging Face token found.", "red"))
print("Run `hf auth login` (or set HF_TOKEN) so the script can reach your private Space.")
sys.exit(1)
def space_host(space_id: str) -> str:
# owner/name -> owner-name.hf.space (HF lowercases and dashes the id)
return space_id.replace("/", "-").lower() + ".hf.space"
# --------------------------------------------------------------------------- #
# scan projects (read the REAL cwd from inside each file — like the engine does)
# --------------------------------------------------------------------------- #
def read_cwd(path: str):
try:
with open(path, "r", encoding="utf-8") as fh:
for line in fh:
line = line.strip()
if not line:
continue
try:
r = json.loads(line)
except ValueError:
continue
if isinstance(r, dict) and r.get("type") in ("user", "assistant") and r.get("cwd"):
return r.get("cwd")
except OSError:
return None
return None
def scan(projects_dir: str):
"""Return [{encoded, cwd, files:[paths]}] grouped by the encoded project folder."""
groups = {}
for fp in glob.glob(os.path.join(projects_dir, "*", "*.jsonl")):
enc = os.path.basename(os.path.dirname(fp))
groups.setdefault(enc, {"encoded": enc, "cwd": None, "files": []})
groups[enc]["files"].append(os.path.abspath(fp))
for g in groups.values():
g["files"].sort()
for f in g["files"]:
cwd = read_cwd(f)
if cwd:
g["cwd"] = cwd
break
out = list(groups.values())
out.sort(key=lambda g: (g["cwd"] or g["encoded"]).lower())
return out
def parse_selection(sel: str, n: int):
sel = sel.strip().lower()
if sel in ("all", "*", "a"):
return list(range(n))
picked = set()
for part in sel.replace(" ", "").split(","):
if not part:
continue
if "-" in part:
try:
a, b = part.split("-", 1)
for i in range(int(a), int(b) + 1):
if 1 <= i <= n:
picked.add(i - 1)
except ValueError:
pass
elif part.isdigit():
i = int(part)
if 1 <= i <= n:
picked.add(i - 1)
return sorted(picked)
# --------------------------------------------------------------------------- #
# scrubber — best-effort secret redaction (you review the summary before upload)
# --------------------------------------------------------------------------- #
_REPL = "[REDACTED]"
_WHOLE = [
("private key block", re.compile(r"-----BEGIN [A-Z ]*PRIVATE KEY-----.*?-----END [A-Z ]*PRIVATE KEY-----", re.S)),
("openai/anthropic key", re.compile(r"\b(?:sk|sk-ant|sk-proj)-[A-Za-z0-9_\-]{20,}\b")),
("hf token", re.compile(r"\bhf_[A-Za-z0-9]{20,}\b")),
("github token", re.compile(r"\bgh[posru]_[A-Za-z0-9]{30,}\b")),
("aws access key id", re.compile(r"\b(?:AKIA|ASIA)[0-9A-Z]{16}\b")),
("google api key", re.compile(r"\bAIza[0-9A-Za-z_\-]{35}\b")),
("slack token", re.compile(r"\bxox[baprs]-[A-Za-z0-9-]{10,}\b")),
("bearer token", re.compile(r"(?i)\bBearer\s+[A-Za-z0-9._\-]{16,}")),
("jwt", re.compile(r"\beyJ[A-Za-z0-9_\-]{10,}\.[A-Za-z0-9_\-]{10,}\.[A-Za-z0-9_\-]{10,}\b")),
]
# group1 = the key + separator (+ an optional opening quote, possibly JSON-escaped as \");
# group2 = the secret value (stops at a quote, backslash, whitespace, or JSON delimiter,
# so it works whether the value is bare or wrapped in escaped quotes inside the JSONL).
_KV = re.compile(
r"(?i)(\"?(?:password|passwd|secret|token|api[_-]?key|access[_-]?key|client[_-]?secret|auth[_-]?token)\"?\s*[:=]\s*(?:\\?\")?)"
r"([^\"\\\s,}{]{6,})"
)
def scrub_text(text: str):
counts = {}
for name, pat in _WHOLE:
text, n = pat.subn(_REPL, text)
if n:
counts[name] = counts.get(name, 0) + n
def _kv(m):
return m.group(1) + _REPL
text, n = _KV.subn(_kv, text)
if n:
counts["key=value secret"] = counts.get("key=value secret", 0) + n
return text, counts
# --------------------------------------------------------------------------- #
# upload (stdlib multipart)
# --------------------------------------------------------------------------- #
def upload_file(host: str, token: str, client: str, project: str, filename: str, data: bytes):
boundary = "----her" + uuid.uuid4().hex
pre = b""
for k, v in (("project", project),):
pre += (f"--{boundary}\r\nContent-Disposition: form-data; name=\"{k}\"\r\n\r\n{v}\r\n").encode()
pre += (
f"--{boundary}\r\nContent-Disposition: form-data; name=\"file\"; filename=\"{filename}\"\r\n"
f"Content-Type: application/jsonl\r\n\r\n"
).encode()
body = pre + data + b"\r\n" + f"--{boundary}--\r\n".encode()
req = urllib.request.Request(
f"https://{host}/api/upload",
data=body,
method="POST",
headers={
"Content-Type": f"multipart/form-data; boundary={boundary}",
"Authorization": f"Bearer {token}",
"X-Her-Client": client,
},
)
with urllib.request.urlopen(req, timeout=120) as resp:
return json.loads(resp.read().decode("utf-8"))
# --------------------------------------------------------------------------- #
# main
# --------------------------------------------------------------------------- #
def main():
ap = argparse.ArgumentParser(description="Bulk-upload Claude Code sessions to your Her Space.")
ap.add_argument("--space", default=os.environ.get("HER_SPACE", DEFAULT_SPACE), help="HF Space id (owner/name)")
ap.add_argument("--host", default=os.environ.get("HER_HOST"), help="override the *.hf.space host")
ap.add_argument("--projects-dir", default=os.path.expanduser("~/.claude/projects"))
ap.add_argument("--staging", default=os.path.abspath("./her-staging"))
args = ap.parse_args()
host = args.host or space_host(args.space)
token = hf_token()
client = uuid.uuid4().hex # this upload's private namespace; the open-URL carries it
print(c("\nHer · हेर — bring your sessions in", "bold"))
print(c(f"Space: {args.space} ({host})", "dim"))
print(c(f"Source: {args.projects_dir}", "dim"))
# ---- STEP 1: SELECT ---------------------------------------------------- #
hr(); print(c("STEP 1 / 3 · choose projects", "cyan"))
groups = scan(args.projects_dir)
if not groups:
print(c(f"No .jsonl sessions found under {args.projects_dir}", "red"))
sys.exit(1)
for i, g in enumerate(groups, 1):
print(f" {i:>2}. {c(g['cwd'] or g['encoded'], 'orange')} "
+ c(f"({len(g['files'])} session{'s' if len(g['files'])!=1 else ''})", "dim"))
print(c("\nEnter numbers (e.g. 1,3,5 or 2-6), or 'all'.", "dim"))
picks = parse_selection(ask("Select projects: "), len(groups))
if not picks:
print("Nothing selected."); sys.exit(0)
chosen = [groups[i] for i in picks]
total_files = sum(len(g["files"]) for g in chosen)
print(c(f"\n→ {len(chosen)} project(s), {total_files} session(s) selected.", "green"))
if not confirm("Copy these into the staging folder and continue?"):
sys.exit(0)
# ---- STEP 2: COPY + SCRUB --------------------------------------------- #
hr(); print(c("STEP 2 / 3 · copy to staging + scrub secrets", "cyan"))
staging = Path(args.staging)
if staging.exists():
shutil.rmtree(staging, ignore_errors=True)
staging.mkdir(parents=True, exist_ok=True)
staged = [] # (project_encoded, staged_path, original_name)
redaction_totals = {}
files_with_redactions = 0
for g in chosen:
outdir = staging / g["encoded"]
outdir.mkdir(parents=True, exist_ok=True)
for src in g["files"]:
try:
raw = Path(src).read_text(encoding="utf-8", errors="replace")
except OSError:
continue
cleaned, counts = scrub_text(raw)
if counts:
files_with_redactions += 1
for k, v in counts.items():
redaction_totals[k] = redaction_totals.get(k, 0) + v
dst = outdir / os.path.basename(src)
dst.write_text(cleaned, encoding="utf-8")
staged.append((g["encoded"], dst, os.path.basename(src)))
print(c(f"Staged {len(staged)} scrubbed session(s) → {staging}", "green"))
if redaction_totals:
print(c(f"Redacted likely secrets in {files_with_redactions} file(s):", "orange"))
for k, v in sorted(redaction_totals.items(), key=lambda x: -x[1]):
print(f" · {k}: {v}")
else:
print(c("No obvious secrets matched (the scrubber is best-effort — review if unsure).", "dim"))
print(c(f"\nYou can inspect the scrubbed copies in {staging} before uploading.", "dim"))
if not confirm("Upload these scrubbed sessions to your private Space?"):
print("Stopped before upload. Staging kept for your review."); sys.exit(0)
# ---- STEP 3: UPLOAD ---------------------------------------------------- #
hr(); print(c("STEP 3 / 3 · upload", "cyan"))
ok = 0
for idx, (enc, path, name) in enumerate(staged, 1):
try:
data = path.read_bytes()
upload_file(host, token, client, enc, name, data)
ok += 1
print(f" [{idx}/{len(staged)}] {c('uploaded', 'green')} {enc}/{name}")
except urllib.error.HTTPError as e:
print(f" [{idx}/{len(staged)}] {c('FAILED', 'red')} {name}: HTTP {e.code} {e.reason}")
except Exception as e: # noqa: BLE001
print(f" [{idx}/{len(staged)}] {c('FAILED', 'red')} {name}: {e}")
hr()
if ok == 0:
print(c("No sessions uploaded.", "red")); sys.exit(1)
print(c(f"✅ Uploaded {ok}/{len(staged)} session(s).", "green"))
spaces_url = f"https://huggingface.co/spaces/{args.space}?client={client}"
print("\nOpen your Projects view (bound to this upload):")
print(" " + c(spaces_url, "cyan"))
print(c("\n⏳ Give it a few seconds on first open — the Space analyzes the sessions and", "orange"))
print(c(" the local model writes the cross-session summary. If a project briefly shows", "orange"))
print(c(" “no sessions found”, just wait a moment and refresh; it’s still generating.", "orange"))
print(c("\nIf your projects don't appear, open the Space, then in the browser console run:", "dim"))
print(c(f" localStorage.setItem('her.clientId','{client}'); location.reload()", "dim"))
print(c("\nReminder: your uploads auto-delete after 24h, or instantly via “clear my data”.", "dim"))
if __name__ == "__main__":
main()
|