fix: Ensure container users always exist

examples/userhandling/tesseract_api.py

@@ -0,0 +1,91 @@
+# Copyright 2025 Pasteur Labs. All Rights Reserved.
+# SPDX-License-Identifier: Apache-2.0
+
+# Tesseract API module for userhandler
+# Generated by tesseract 1.2.1.dev14+g667850bb1.d20251217 on 2025-12-18T14:34:34.613891
+
+
+import os
+import pwd
+
+from pydantic import BaseModel
+
+#
+# Schemas
+#
+
+
+class InputSchema(BaseModel):
+    pass
+
+
+class OutputSchema(BaseModel):
+    uid: int
+    gid: int
+    username: str
+    home: str
+    shell: str
+
+
+#
+# Required endpoints
+#
+
+
+def apply(inputs: InputSchema) -> OutputSchema:
+    """A simple Tesseract that fails if the current OS user doesn't have a valid /etc/passwd entry."""
+    # Get current user ID
+    uid = os.getuid()
+    gid = os.getgid()
+
+    # Try to get the passwd entry for the current user
+    try:
+        pw_entry = pwd.getpwuid(uid)
+    except KeyError as exc:
+        raise RuntimeError(
+            f"Current user (UID: {uid}) does not have an entry in /etc/passwd. "
+            "This container should have run addmeplease in the entrypoint."
+        ) from exc
+
+    if pw_entry.pw_gid != gid:
+        raise RuntimeError(
+            f"Current group (GID: {gid}) does not match expected group from /etc/passwd ({pw_entry.pw_gid})."
+        )
+
+    # Return the user information
+    return OutputSchema(
+        uid=pw_entry.pw_uid,
+        gid=pw_entry.pw_gid,
+        username=pw_entry.pw_name,
+        home=pw_entry.pw_dir,
+        shell=pw_entry.pw_shell,
+    )
+
+
+#
+# Optional endpoints
+#
+
+# import numpy as np
+
+# def jacobian(inputs: InputSchema, jac_inputs: set[str], jac_outputs: set[str]):
+#     return {}
+
+# def jacobian_vector_product(
+#     inputs: InputSchema,
+#     jvp_inputs: set[str],
+#     jvp_outputs: set[str],
+#     tangent_vector: dict[str, np.typing.ArrayLike]
+# ) -> dict[str, np.typing.ArrayLike]:
+#     return {}
+
+# def vector_jacobian_product(
+#     inputs: InputSchema,
+#     vjp_inputs: set[str],
+#     vjp_outputs: set[str],
+#     cotangent_vector: dict[str, np.typing.ArrayLike]
+# ) -> dict[str, np.typing.ArrayLike]:
+#     return {}
+
+# def abstract_eval(abstract_inputs):
+#     return {}

examples/userhandling/tesseract_config.yaml

@@ -0,0 +1,27 @@
+# Tesseract configuration file
+# Generated by tesseract 1.2.1.dev14+g667850bb1.d20251217 on 2025-12-18T14:34:34.613891
+
+name: "userhandler"
+version: "unknown"
+description: ""
+
+build_config:
+  # Base image to use for the container, must be Ubuntu or Debian-based
+  # base_image: "debian:bookworm-slim"
+
+  # Platform to build the container for. In general, images can only be executed
+  # on the platform they were built for.
+  target_platform: "native"
+
+  # Additional packages to install in the container (via apt-get)
+  # extra_packages:
+  #  - package_name
+
+  # Data to copy into the container, relative to the project root
+  # package_data:
+  #  - [path/to/source, path/to/destination]
+
+  # Additional Dockerfile commands to run during the build process
+  # custom_build_steps:
+  #  - |
+  #   RUN echo "Hello, World!"

examples/userhandling/tesseract_requirements.txt

@@ -0,0 +1,9 @@
+# Tesseract requirements file
+# Generated by tesseract 1.2.1.dev14+g667850bb1.d20251217 on 2025-12-18T14:34:34.613891
+
+# Add Python requirements like this:
+# numpy==1.18.1
+
+# This may contain private dependencies via SSH URLs:
+# git+ssh://git@github.com/username/repo.git@branch
+# (use `tesseract build --forward-ssh-agent` to grant the builder access to your SSH keys)

tesseract_core/sdk/engine.py

@@ -214,11 +214,13 @@ def prepare_build_context(
 
     template_dir = get_template_dir()
 
+    extra_files = [template_dir / "entrypoint.sh", template_dir / "addmeplease.c"]
+
     requirement_config = user_config.build_config.requirements
-    copy(
-        template_dir / requirement_config._build_script,
-        context_dir / "__tesseract_source__" / requirement_config._build_script,
-    )
+    extra_files.append(template_dir / requirement_config._build_script)
+
+    for path in extra_files:
+        copy(path, context_dir / path.relative_to(template_dir))
 
     # When building from a requirements.txt we support local dependencies.
     # We separate local dep. lines from the requirements.txt and copy the

tesseract_core/sdk/templates/Dockerfile.base

@@ -1,3 +1,6 @@
+# Copyright 2025 Pasteur Labs. All Rights Reserved.
+# SPDX-License-Identifier: Apache-2.0
+
 # Stage 1: Build Python environment and install requirements
 {% if config.build_config.target_platform.strip() == "native" %}
 FROM {{ config.build_config.base_image }} AS build_stage
@@ -22,8 +25,19 @@ fi
 RUN apt-get update && apt-get install -y --no-install-recommends \
         git \
         ssh \
+        # Needed to compile addmeplease, and nice to have in general
+        gcc \
+        libc6-dev \
     && rm -rf /var/lib/apt/lists/*
 
+# Copy and compile addmeplease
+COPY "addmeplease.c" /tmp/addmeplease.c
+RUN gcc -o /bin/addmeplease /tmp/addmeplease.c && \
+    rm /tmp/addmeplease.c && \
+    chown root:root /bin/addmeplease && \
+    chmod 4755 /bin/addmeplease && \
+    ls -l /bin/addmeplease
+
 {% if config.build_config.extra_packages %}
 # Install extra packages
 RUN apt-get update && apt-get install -y --no-install-recommends \
@@ -40,7 +54,7 @@ RUN mkdir -p -m 0700 ~/.ssh && ssh-keyscan github.com >> ~/.ssh/known_hosts
 WORKDIR /tmp/build
 COPY {{ tesseract_runtime_location }} ./tesseract_runtime/
 COPY {{ tesseract_source_directory }}/{{ config.build_config.requirements._filename }} ./
-COPY {{ tesseract_source_directory }}/{{ config.build_config.requirements._build_script }} ./
+COPY {{ config.build_config.requirements._build_script }} ./
 COPY local_requirements/ ./local_requirements
 
 # Build a python venv from python provider build scripts.
@@ -79,6 +93,9 @@ ENV TESSERACT_NAME="{{ config.name | replace('"', '\\"') | replace('\n', '\\n')
 
 # Copy only necessary files
 COPY --from=build_stage /python-env /python-env
+COPY --from=build_stage /bin/addmeplease /bin/addmeplease
+COPY "entrypoint.sh" /tesseract/entrypoint.sh
+RUN chmod 555 /tesseract/entrypoint.sh
 COPY "{{ tesseract_source_directory }}/tesseract_api.py" ${TESSERACT_API_PATH}
 RUN chmod 444 ${TESSERACT_API_PATH}
 
@@ -102,14 +119,17 @@ RUN mkdir -p /tesseract/input_data /tesseract/output_data && \
     chmod 777 /tesseract/output_data
 VOLUME ["/tesseract/input_data", "/tesseract/output_data"]
 
-ENV HOME=/tesseract
 ENV TESSERACT_INPUT_PATH="/tesseract/input_data"
 ENV TESSERACT_OUTPUT_PATH="/tesseract/output_data"
 
+# Drop to random non-root user to prevent exploits
+# (this way, we never start the container as root unless specifically requested)
+USER 9999:9999
+
 # Final sanity check to ensure the runtime is installed and tesseract_api.py is valid
 {% if not config.build_config.skip_checks %}
-RUN _TESSERACT_IS_BUILDING=1 tesseract-runtime check
+RUN _TESSERACT_IS_BUILDING=1 /tesseract/entrypoint.sh tesseract-runtime check
 {% endif %}
 
-ENTRYPOINT ["tesseract-runtime"]
+ENTRYPOINT ["/tesseract/entrypoint.sh", "tesseract-runtime"]
 CMD ["--help"]

tesseract_core/sdk/templates/addmeplease.c

@@ -0,0 +1,168 @@
+// Copyright 2025 Pasteur Labs. All Rights Reserved.
+// SPDX-License-Identifier: Apache-2.0
+
+/*
+ * addmeplease - Create user/group entries for non-privileged container users
+ *
+ * This program allows a non-privileged user to create their own entry in
+ * /etc/passwd and /etc/group. This is necessary because:
+ *
+ * 1. Containers run as non-root users (e.g., USER 501:501) for security
+ * 2. Some applications require an entry in /etc/passwd to function properly
+ * 3. The container entrypoint runs as the non-privileged user and cannot
+ *    directly modify /etc/passwd without elevated privileges
+ *
+ * This binary is compiled and installed with setuid root (chmod 4755), which
+ * allows it to temporarily elevate privileges to modify /etc/passwd and
+ * /etc/group, then immediately drop privileges back to the calling user.
+ *
+ * Note: setuid only works with compiled binaries, not shell scripts. This is
+ * why this must be implemented in C rather than as a bash script.
+ */
+
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <pwd.h>
+#include <grp.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <sys/stat.h>
+
+#define PASSWD_FILE "/etc/passwd"
+#define GROUP_FILE "/etc/group"
+#define SHADOW_FILE "/etc/shadow"
+#define GSHADOW_FILE "/etc/gshadow"
+
+int group_exists(gid_t gid) {
+    struct group *grp = getgrgid(gid);
+    return (grp != NULL);
+}
+
+int user_exists(uid_t uid) {
+    struct passwd *pwd = getpwuid(uid);
+    return (pwd != NULL);
+}
+
+int append_to_file(const char *filepath, const char *line) {
+    int fd = open(filepath, O_WRONLY | O_APPEND | O_CREAT, 0644);
+    if (fd < 0) {
+        return -1;
+    }
+
+    size_t len = strlen(line);
+    ssize_t written = write(fd, line, len);
+    close(fd);
+
+    return (written == (ssize_t)len) ? 0 : -1;
+}
+
+int create_group(gid_t gid, const char *groupname) {
+    if (group_exists(gid)) {
+        return 0;
+    }
+
+    // Format: groupname:x:gid:
+    char group_line[256];
+    snprintf(group_line, sizeof(group_line), "%s:x:%d:\n", groupname, gid);
+
+    if (append_to_file(GROUP_FILE, group_line) != 0) {
+        fprintf(stderr, "addmeplease: Failed to add group to %s: %s\n",
+                GROUP_FILE, strerror(errno));
+        return -1;
+    }
+
+    // Add to gshadow if it exists
+    char gshadow_line[256];
+    snprintf(gshadow_line, sizeof(gshadow_line), "%s:!::\n", groupname);
+
+    struct stat st;
+    if (stat(GSHADOW_FILE, &st) == 0) {
+        append_to_file(GSHADOW_FILE, gshadow_line);
+    }
+
+    return 0;
+}
+
+int create_user(uid_t uid, gid_t gid, const char *username) {
+    if (user_exists(uid)) {
+        return 0;
+    }
+
+    // Format: username:x:uid:gid:comment:home:shell
+    char passwd_line[512];
+    snprintf(passwd_line, sizeof(passwd_line),
+             "%s:x:%d:%d:Tesseract User:/tesseract:/bin/bash\n",
+             username, uid, gid);
+
+    if (append_to_file(PASSWD_FILE, passwd_line) != 0) {
+        fprintf(stderr, "addmeplease: Failed to add user to %s: %s\n",
+                PASSWD_FILE, strerror(errno));
+        return -1;
+    }
+
+    // Add to shadow file with locked password
+    // Format: username:!:lastchanged:min:max:warn:inactive:expire:
+    // We use ! for locked password and 0 for lastchanged (epoch)
+    char shadow_line[256];
+    snprintf(shadow_line, sizeof(shadow_line), "%s:!:0:0:99999:7:::\n", username);
+
+    struct stat st;
+    if (stat(SHADOW_FILE, &st) == 0) {
+        // Shadow file typically has restricted permissions
+        int fd = open(SHADOW_FILE, O_WRONLY | O_APPEND);
+        if (fd >= 0) {
+            size_t len = strlen(shadow_line);
+            write(fd, shadow_line, len);
+            close(fd);
+        }
+    }
+
+    return 0;
+}
+
+int main(void) {
+    uid_t real_uid = getuid();
+    gid_t real_gid = getgid();
+
+    // If running as root, nothing to do
+    if (real_uid == 0) {
+        return 0;
+    }
+
+    // Check if user already exists
+    if (user_exists(real_uid)) {
+        return 0;
+    }
+
+    // We need root privileges to modify /etc/passwd and /etc/group
+    // This is why the binary needs setuid root
+    if (seteuid(0) != 0) {
+        fprintf(stderr, "addmeplease: Failed to elevate privileges: %s\n", strerror(errno));
+        fprintf(stderr, "addmeplease: This binary must be setuid root (chmod 4755)\n");
+        return 1;
+    }
+
+    // Create group if it doesn't exist
+    if (create_group(real_gid, "tesseract-group") != 0) {
+        fprintf(stderr, "addmeplease: Warning: Failed to create group\n");
+        // Continue anyway, the GID might exist with a different name
+    }
+
+    // Create user with current UID/GID
+    if (create_user(real_uid, real_gid, "tesseract-user") != 0) {
+        fprintf(stderr, "addmeplease: Warning: Failed to create user\n");
+        // Continue anyway
+    }
+
+    // Drop privileges back to the original user
+    if (seteuid(real_uid) != 0) {
+        fprintf(stderr, "addmeplease: Failed to drop privileges: %s\n", strerror(errno));
+        return 1;
+    }
+
+    return 0;
+}

tesseract_core/sdk/templates/build_conda_venv.sh

@@ -1,5 +1,8 @@
 #!/bin/bash
 
+# Copyright 2025 Pasteur Labs. All Rights Reserved.
+# SPDX-License-Identifier: Apache-2.0
+
 set -e  # Exit immediately if a command exits with a non-zero status
 
 conda env create --file tesseract_environment.yaml -p /python-env --quiet