# eling-agent - Doramagic AI Context Pack

> Positioning: a pre-install experience and judgment asset. It helps the host AI get off to a good start, but it does not mean the project has already been installed, run, or validated.

## Sufficiency Principle

- **Sufficiency over compression**: The AI Context Pack should be sufficient for the host AI to understand the project's value, capability boundaries, entrypoints, risks, and evidence sources before starting work; it may be layered, but it does not aim for the shortest possible summary.
- **Compression policy**: Compress only noise and duplication, never context that affects judgment or the quality of the work.

## How the Host AI Should Use This

You are reading the AI Context Pack that Doramagic compiled for eling-agent. Treat it as pre-work context: help the user understand who it fits, what it can do, how to start, what must be verified after install, and where the risks are. Do not claim that you have already installed, run, or executed the target project.

## Claim Consumption Rules

- **Fact source**: Repo Evidence + Claim/Evidence Graph; the Human Wiki only supplies salience, terminology, and narrative structure.
- **Minimum status for a fact**: `supported`
- `supported`: May be used as a project fact, but the answer must cite the claim_id and evidence path.
- `weak`: Usable only as a low-confidence lead; the user must be asked to keep verifying.
- `inferred`: Usable only for risk notes or open questions; must not be packaged as a project fact.
- `unverified`: Must not be used as fact; state clearly that evidence is insufficient.
- `contradicted`: Must show the conflicting sources and must not force a single version on the user's behalf.

## Who It Fits Best

- **Developers already using host AIs such as Claude/Codex/Cursor/Gemini**: The README or plugin config mentions multiple host AIs. Evidence: `README.md` Claim: `clm_0003` supported 0.86
- **Users who want to bring professional workflows into a host AI**: The repo contains Skill documents. Evidence: `src/eling/zero_plugin/SKILL.md` Claim: `clm_0004` supported 0.86

## What It Can Do

- **AI Skill / Agent Instruction Asset Library** (Previewable before install): The project contains Skill or Agent instruction files that a host AI can read, useful for bringing professional workflows into hosts like Claude, Codex, or Cursor. Evidence: `src/eling/zero_plugin/SKILL.md` Claim: `clm_0001` supported 0.86
- **Command-Line Startup or Install Flow** (Verify after install): The project documentation contains runnable commands; real use requires running them in a local or host environment. Evidence: `README.md` Claim: `clm_0002` supported 0.86

## How to Start

- `git clone https://github.com/PatrickNoFilter/eling-agent.git` Evidence: `README.md` Claim: `clm_0005` supported 0.86
- `pip install -r requirements.txt` Evidence: `README.md` Claim: `clm_0006` supported 0.86
- `pip install "eling-agent[hrr,embeddings,notion,markdownify]"` Evidence: `README.md` Claim: `clm_0007` supported 0.86

## Continue-or-Stop Decision Card

- **Current recommendation**: Sandbox-trial permissions first
- **Why**: The project has signals of install commands, host configuration, or local writes; do not go straight into your primary environment—trial it in isolation first.

### 30-Second Read

- **What to do now**: Sandbox-trial permissions first
- **Minimum safe next step**: Run Prompt Preview first; if you still want to install, trial only in an isolated environment
- **Do not trust yet**: Tool permission boundaries cannot be trusted before install.
- **Continuing will touch**: Command execution, Host AI configuration, Local environment or project files

### What You Can Trust Now

- **Target-audience signal: Developers already using host AIs such as Claude/Codex/Cursor/Gemini** (supported): Backed by a supported claim or project evidence, but that still is not the same as real install results. Evidence: `README.md` Claim: `clm_0003` supported 0.86
- **Target-audience signal: Users who want to bring professional workflows into a host AI** (supported): Backed by a supported claim or project evidence, but that still is not the same as real install results. Evidence: `src/eling/zero_plugin/SKILL.md` Claim: `clm_0004` supported 0.86
- **Capability exists: AI Skill / Agent Instruction Asset Library** (supported): You can trust that the project contains signals of this capability; whether it fits your specific task still needs trial or after-install verification. Evidence: `src/eling/zero_plugin/SKILL.md` Claim: `clm_0001` supported 0.86
- **Capability exists: Command-Line Startup or Install Flow** (supported): You can trust that the project contains signals of this capability; whether it fits your specific task still needs trial or after-install verification. Evidence: `README.md` Claim: `clm_0002` supported 0.86
- **There are Quick Start / install-command signals** (supported): You can trust that the docs mention a startup or install entrypoint; do not run it directly in your primary environment because of that. Evidence: `README.md` Claim: `clm_0005` supported 0.86

### What You Cannot Trust Yet

- **Tool permission boundaries cannot be trusted before install.** (unverified): MCP/tool projects usually touch files, the network, the browser, or external APIs, so permissions and logs must be checked for real.
- **Real output quality cannot be trusted before install.** (unverified): Prompt Preview can only show how it guides you; it cannot prove result quality in the real project.
- **Host AI version compatibility cannot be trusted before install.** (unverified): Host loading rules and version differences across Claude, Cursor, Codex, Gemini, and others must be verified in a real environment.
- **That it will not pollute your existing host AI's behavior cannot be trusted directly.** (inferred): Skill, plugin, and AGENTS/CLAUDE/GEMINI instructions may change the host AI's default behavior. Evidence: `src/eling/zero_plugin/SKILL.md`
- **Safe rollback cannot be assumed by default.** (unverified): Unless the project clearly provides uninstall and recovery instructions, verify in an isolated environment first.
- **After a real install, is it compatible with the user's current host AI version?** (unverified): Compatibility can only be verified in the actual host environment.
- **Does the project's output quality meet the user's specific task?** (unverified): The pre-install preview can only show flow and boundaries; it cannot replace real evaluation.
- **Do the install commands require network access, permissions, or global writes?** (unverified): This affects install risk in both enterprise and personal environments. Evidence: `README.md`

### What Continuing Will Touch

- **Command execution**: Package managers, network downloads, the local plugin directory, project config, or the user's home directory. Why: Running the very first command can already change your environment; decide whether it is worth running first. Evidence: `README.md`
- **Host AI configuration**: The plugin, Skill, or rule-loading config of hosts like Claude/Codex/Cursor/Gemini/OpenCode. Why: Host configuration changes how the AI works afterward and may conflict with the user's existing rules. Evidence: `src/eling/zero_plugin/SKILL.md`
- **Local environment or project files**: Install results, plugin caches, project config, or local dependency directories. Why: The write scope and rollback path cannot be proven before install and need isolated verification. Evidence: `README.md`
- **Host AI context**: The AI Context Pack, Prompt Preview, Skill routing, risk rules, and project facts. Why: Importing context affects the host AI's later judgment, so avoid packaging unverified items as facts.

### Minimum Safe Next Steps

- **Run Prompt Preview first**: Use a pre-install interactive trial to judge whether the way of working fits; it needs no authorization or environment change. (applies when: Applies to any project, especially when output quality is unknown.)
- **Trial-install only in an isolated directory or a test account**: Avoid letting install commands pollute your primary host AI, real projects, or home directory. (applies when: When there are signals of command execution, plugin config, or local writes.)
- **Back up your host AI configuration first**: Skill, plugin, and rule files may change the default behavior of Claude/Cursor/Codex. (applies when: When there is a plugin manifest, a Skill, or a host rule entrypoint.)
- **After install, verify just one minimal task**: Verify loading, compatibility, output quality, and rollback first, then decide whether to use it deeply. (applies when: When moving from a trial into a real workflow.)

### Exit Plan

- **Preserve the pre-install state**: Record the original host config and project state so you can later judge whether it is recoverable.
- **Be ready to remove the host plugin / Skill / rule entrypoint**: If behavior is off after the trial install, you can restore the host AI to its pre-trial state.
- **Record the install commands and written paths**: Without clear uninstall instructions, you at least need to know which directories or configs to clean up manually.
- **If there is no rollback path, do not enter your primary environment**: No rollback is a blocker before continuing; do not proceed on trust or luck.

## What Can Only Be Previewed

- Explain who the project fits and what it can do
- Demonstrate a typical conversation flow based on project docs
- Help the user decide whether it is worth installing or researching further

## What Must Be Verified After Install

- Actually installing the Skill, plugin, or CLI
- Running scripts, modifying local files, or accessing external services
- Verifying real output quality, performance, and compatibility

## Boundary & Risk Decision Card

- **Mistaking the pre-install preview for a real run**: The user may overestimate how much configuration, permission, and compatibility verification the project has already done. Mitigation: Clearly separate prompt_preview_can_do from runtime_required. Claim: `clm_0008` inferred 0.45
- **Command execution will modify the local environment**: Install commands may write to the user's home directory, the host plugin directory, or project configuration. Mitigation: Run in an isolated environment or a test account first. Evidence: `README.md` Claim: `clm_0009` supported 0.86
- **To confirm**: After a real install, is it compatible with the user's current host AI version?. Why: Compatibility can only be verified in the actual host environment.
- **To confirm**: Does the project's output quality meet the user's specific task?. Why: The pre-install preview can only show flow and boundaries; it cannot replace real evaluation.
- **To confirm**: Do the install commands require network access, permissions, or global writes?. Why: This affects install risk in both enterprise and personal environments.

## Pre-Work Working Context

### Loading Order

- First read how_to_use.host_ai_instruction to establish the boundaries of this pre-install judgment asset.
- Read claim_graph_summary to confirm facts come from the Claim/Evidence Graph, not the Human Wiki narrative.
- Then read intended_users, capabilities, and quick_start_candidates to judge whether the user is a match.
- When you need to carry out a concrete task, check role_skill_index first, then evidence_index.
- For real install, file modification, network access, performance, or compatibility questions, turn to risk_card and boundaries.runtime_required.

### Task Routes

- **AI Skill / Agent Instruction Asset Library**: Use role_skill_index / evidence_index to help the user pick a usable role, Skill, or workflow first. Boundary: Can be experienced via a pre-install Prompt. Evidence: `src/eling/zero_plugin/SKILL.md` Claim: `clm_0001` supported 0.86
- **Command-Line Startup or Install Flow**: State that this is an after-install capability first, then give a pre-install checklist. Boundary: Must be verified after a real install or run. Evidence: `README.md` Claim: `clm_0002` supported 0.86

### Context Scale

- Total files: 86
- Important-file coverage: 40/86
- Evidence index entries: 53
- Role / Skill entries: 1

### Handling Insufficient Evidence

- **missing_evidence**: State that evidence is insufficient and ask the user for the target file, a README section, or after-install verification records; do not fill in facts.
- **out_of_scope_request**: State that the task is beyond the current AI Context Pack's evidence scope and suggest the user check the Human Manual or verify after a real install.
- **runtime_request**: Provide a pre-install checklist and command sources, but do not run commands for the user or claim they have been run.
- **source_conflict**: Show the conflicting sources side by side, mark them as unverified, and do not force a single version.

## Prompt Recipes

### Fit assessment

- Goal: Judge whether this project fits the user's current task.
- Expected output: A fit conclusion, key reasons, evidence citations, what can be previewed before install, what must be verified after install, and a next-step recommendation.

```text
Based on the AI Context Pack for eling-agent, ask me 3 necessary questions first, then judge whether it fits my task. The answer must cover: who it fits, what it can do, what it cannot do, whether it is worth installing, and where the evidence comes from. Every project fact must cite evidence_refs, source_paths, or a claim_id.
```

### Pre-install experience

- Goal: Let the user feel the core workflow before installing, while avoiding packaging the preview as real capability or a marketing promise.
- Expected output: An experience script with boundary labels, an after-install verification checklist, and a cautious recommendation; with no real-run promises or strong marketing language.

```text
Treat eling-agent as a pre-install experience asset, not an already-installed tool or a real runtime environment.

Output exactly four parts:
1. Ask me 3 necessary questions first.
2. Give an "experience script": use the three labels [Previewable before install], [Must verify after install], and [Insufficient evidence] to show how it might guide the workflow.
3. Give an after-install verification checklist: list which capabilities can only be confirmed after a real install, real host loading, and a real project run.
4. Give a cautious recommendation: only "worth researching/trialing further", "add information before deciding", or "not recommended to continue"; do not endorse the project.

Hard boundaries:
- Do not claim you have installed, run, executed tests, modified files, or produced real results.
- Do not write promise-like phrasing such as "auto-adapts", "guarantees passing", "perfect fit", or "strongly recommend installing".
- If you describe how it works after install, you must use a conditional such as "if installed successfully and the host loads the Skill correctly, it might...".
- The experience script may only be written as "example lines / hypothetical flow": use "might ask / might suggest / might show", not "has written, has generated, has passed, is running, is generating".
- Prompt Preview does not hand out install commands; if the user is ready to trial, only prompt them to read Quick Start and the Risk Card first and to verify in an isolated environment.
- Every project fact must come from a supported claim, evidence_refs, or source_paths; inferred/unverified items can only be risks or open questions.

```

### Role / Skill selection

- Goal: Pick the best-matching asset from the project's roles or Skills.
- Expected output: A list of candidate roles or Skills, each with an applicable scenario, evidence paths, risk boundary, and whether after-install verification is needed.

```text
Read role_skill_index and recommend 3-5 of the most relevant roles or Skills for my target task. For each recommendation, state the applicable scenario, likely output, risk boundary, and evidence_refs.
```

### Risk pre-check

- Goal: Identify environment, permission, rule-conflict, and quality risks before installing or adopting.
- Expected output: A checklist of environment, permission, dependency, license, host-conflict, quality risk, and unknown items.

```text
Based on risk_card, boundaries, and quick_start_candidates, give me a pre-install risk pre-check list. Do not run commands for me; only explain what I should check, why, and what impact a failure would have.
```

### Host AI kickoff instruction

- Goal: Turn the project context into a host AI instruction for the start of a conversation.
- Expected output: A pre-work instruction with clear boundaries and clear evidence citations, suitable to copy to a host AI.

```text
Based on the AI Context Pack for eling-agent, generate a pre-work instruction I can paste to my host AI. This instruction must obey not_runtime=true and must not claim the project has been installed, run, or produced real results.
```

## Role / Skill Index

- Indexed 1 role / Skill / project-doc entries.

- **Eling Memory System MCP Split** (skill): Eling has been split into two MCP servers : Activation hint: When the user's task is highly relevant to the workflow described by “Eling Memory System MCP Split”, use it for a pre-install experience first, then decide whether to install. Evidence: `src/eling/zero_plugin/SKILL.md`

## Evidence Index

- Indexed 53 evidence entries.

- **🧠 Eling Agent** (documentation): Personal auto-learning agent CLI — memory, skills, MCP tools, workspace, and terminal UI Evidence: `README.md`
- **Eling Memory System MCP Split** (skill_instruction): Eling has been split into two MCP servers : Evidence: `src/eling/zero_plugin/SKILL.md`
- **Architecture** (documentation): Eling Agent follows a modular, event-driven architecture: Evidence: `docs/architecture.md`
- **Configuration** (documentation): See config.example.json ../config.example.json for the template. Evidence: `docs/configuration.md`
- **Eling Agent Documentation** (documentation): Eling Agent is a personal auto-learning agent CLI that runs in your terminal. Evidence: `docs/index.md`
- **License** (source_file): Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files the "Software" , to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: Evidence: `LICENSE`
- **MCP Integration** (documentation): Eling Agent supports the Model Context Protocol MCP for connecting to external tools. Evidence: `docs/mcp.md`
- **Memory System** (documentation): Eling Agent uses a multi-layer memory architecture: Evidence: `docs/memory.md`
- **Plugins** (documentation): Eling Agent supports a simple plugin system. Evidence: `docs/plugins.md`
- **Config.Example** (structured_config): { "zen api key": "REPLACE WITH YOUR ZEN KEY", "zen base url": "https://opencode.ai/zen/v1", "zen model": "deepseek-v4-flash-free", "max tool rounds": 6, "memory db": "agent memory.db", "skills db": "agent skills.db", "mcp servers": {} } Evidence: `config.example.json`
- **Auto-fix with safe fixes** (source_file): log = logging.getLogger "eling" ⋮---- BASE SYSTEM PROMPT = ⋮---- def load config path: str = "config.json" - dict ⋮---- script dir = os.path.dirname os.path.abspath file path = os.path.join script dir, "config.json" ⋮---- home cfg = os.path.join os.path.expanduser "~" , "eling-agent", "config.json" ⋮---- path = home cfg ⋮---- example = path.replace "config.json", "config.example.json" ⋮---- cfg = json.load f ⋮---- parts = BASE SYSTEM PROMPT ⋮---- PY FILE RE = re.compile r' ?:^ \s / ^\s \.py a-zA-Z0-9 ./- +\.py ' ⋮---- def extract py files tool results: list dict - set str ⋮---- """Extract .py file paths from tool result content.""" files: set str = set ⋮---- text = tr.get "content", "" ⋮---… Evidence: `agent.py`
- **Notify initialized** (source_file): log = logging.getLogger "mcp client" ⋮---- class MCPServerConnection ⋮---- def init self, name: str, command: list str , env: dict = None ⋮---- def next id self - int ⋮---- def start self ⋮---- init resp = self. send request ⋮---- Notify initialized ⋮---- tools resp = self. send request "tools/list", {} ⋮---- def reader self ⋮---- line = line.strip ⋮---- msg = json.loads line ⋮---- def stderr reader self ⋮---- buffer: list str = ⋮---- raw = line.rstrip '\n\r' ⋮---- decorative = len BOX.findall raw ⋮---- text = '; '.join buffer ⋮---- req id = self. next id payload = { ⋮---- timeout = 30 deadline = threading.Timer timeout, lambda: None ⋮---- resp = self. response queue.get timeout=timeout ⋮--… Evidence: `mcp_client.py`
- **Compare against newer entries only avoids double-deletion** (source_file): RECENCY ALPHA = 0.20 ⋮---- RECENCY HALF DAYS = 7.0 ⋮---- class MemoryEntry NamedTuple ⋮---- id: int timestamp: str user input: str agent output: str outcome: str ⋮---- class MemoryStore ⋮---- def init self, db path: str ⋮---- def create table self ⋮---- def hash self, user input: str, agent output: str - str ⋮---- raw = f"{user input.strip } {agent output.strip }" ⋮---- def add self, user input: str, agent output: str, outcome: str = "neutral" - int ⋮---- ts = datetime.now timezone.utc .isoformat content hash = self. hash user input, agent output ⋮---- existing = self. conn.execute ⋮---- cur = self. conn.execute ⋮---- def count self - int ⋮---- row = self. conn.execute "SELECT COUNT AS cnt… Evidence: `memory.py`
- **Init** (source_file): log = logging.getLogger "plugins" ⋮---- PYTHON TO JSON = { ⋮---- def type to schema tp - dict ⋮---- origin = getattr tp, " origin ", None ⋮---- item tp = getattr tp, " args ", str, 0 ⋮---- val tp = getattr tp, " args ", str, 1 ⋮---- def tool name: str, description: str = "", auto schema: bool = True ⋮---- Returns: Evidence: `plugins/__init__.py`
- **── blackbox tools ─────────────────────────────────────────────────** (source_file): log = logging.getLogger "eling integration" ⋮---- brain: Any = None continuum: Any = None blackbox: Any = None markitdown: Any = None ⋮---- def get brain ⋮---- home = os.environ.get "ELING HOME" brain = Brain home=Path home if home else None ⋮---- def get continuum ⋮---- continuum = ContinuumStore ⋮---- def get blackbox ⋮---- blackbox = BlackboxStore ⋮---- def get markitdown ⋮---- markitdown = MarkItDown ⋮---- def empty tool name: str, fn, desc: str - tuple str, dict ⋮---- result = get brain .remember content=content, layer=layer, category=category, ⋮---- result = get brain .recall query=query, layers=layers, limit=limit, ⋮---- def brain probe entity: str, limit: int = 10 - str ⋮---- result… Evidence: `plugins/eling_integration.py`
- **Shell Plugin** (source_file): def run shell command: str - str ⋮---- result = subprocess.run output = result.stdout ⋮---- output = result.stderr ⋮---- output = f" exit code {result.returncode} \n{output}" ⋮---- def exec shell cmd: str - str ⋮---- """Internal worker that runs one command and captures output.""" ⋮---- out = result.stdout ⋮---- out = result.stderr ⋮---- out = f" exit code {result.returncode} \n{out}" ⋮---- def run parallel commands: list str - str ⋮---- """Run multiple shell commands concurrently using thread-level parallelism. Each command spawns its own subprocess GIL-safe , so threads are efficient.""" max workers = os.cpu count or 4 ⋮---- futures = {executor.submit exec shell, cmd : cmd for cmd in comm… Evidence: `plugins/shell_plugin.py`
- **Catch-all for other request failures** (source_file): class ZenProvider ⋮---- payload = { ⋮---- last error = None ⋮---- resp = requests.post ⋮---- wait = min 2 attempt + 1, 60 ⋮---- last error = exc ⋮---- Catch-all for other request failures ⋮---- last error = resp ⋮---- data = resp.json ⋮---- last status = getattr last error, 'status code', None if last error else None Evidence: `provider.py`
- **Pyproject** (source_file): build-system requires = "setuptools =68.0" build-backend = "setuptools.build meta" Evidence: `pyproject.toml`
- **Requirements** (source_file): requests =2.31.0 Evidence: `requirements.txt`
- **── Helpers ────────────────────────────────────────────────────** (source_file): log = logging.getLogger "eling.skills" ⋮---- SENTENCE TRANSFORMERS AVAILABLE = False sentence model = None ⋮---- SENTENCE TRANSFORMERS AVAILABLE = True ⋮---- def get sentence model ⋮---- sentence model = sentence transformers.SentenceTransformer ⋮---- def cosine sim vec a, b - float ⋮---- dot = float np.dot a, b na = float np.linalg.norm a nb = float np.linalg.norm b ⋮---- SUCCESS ALPHA = 0.30 DEDUP THRESHOLD = 0.85 LOW PERFORMER MIN USES = 5 LOW PERFORMER MAX RATE = 0.30 BODY MAX LENGTH = 2000 ⋮---- class Skill NamedTuple ⋮---- id: int name: str trigger: str body: str uses: int successes: int prior body: str None prior trigger: str None created at: str updated at: str ⋮---- class SkillLibr… Evidence: `skills.py`
- **Intersection dot product** (source_file): def tokenize text: str - list str ⋮---- text = str text ⋮---- def vectorize text: str - collections.Counter ⋮---- """Convert text to a Bag-of-Words Counter.""" ⋮---- def cosine sim vec a: collections.Counter, vec b: collections.Counter - float ⋮---- """Compute cosine similarity between two Counters no numpy .""" Intersection dot product dot product = 0 ⋮---- norm a = math.sqrt sum v v for v in vec a.values norm b = math.sqrt sum v v for v in vec b.values ⋮---- """ Return the top-k candidates sorted by cosine similarity descending. Args: query: The search query string. candidates: Iterable of items to score. text fn: Callable item - str, returns the text to compare. k: Maximum number of resu… Evidence: `textsim.py`
- **── Assistant Response spacious, single-border ────────────────** (source_file): def format time seconds: float - str ⋮---- dur = int seconds ⋮---- ROOT = Path file .resolve .parent ⋮---- def get version - str ⋮---- ver file = ROOT / ".agent-version" ⋮---- v = ver file.read text .strip ⋮---- pyproject = ROOT / "pyproject.toml" ⋮---- VERSION = get version ⋮---- THEMES: dict str, dict str, str = { ⋮---- DEFAULT THEME = "cobalt" ⋮---- THEME NAMES = list THEMES.keys ⋮---- def resolve theme name name: str None - str ⋮---- def resolve theme name: str None - dict str, str ⋮---- ACTIVE THEME = resolve theme DEFAULT THEME ACCENT = ACTIVE THEME "ACCENT" MIDBLUE = ACTIVE THEME "MIDBLUE" LIGHTBLUE = ACTIVE THEME "LIGHTBLUE" MUTEDBLUE = ACTIVE THEME "MUTEDBLUE" DEEPBLUE = ACTIVE THE… Evidence: `tui.py`
- **Init** (source_file): version = "0.12.1" all = Evidence: `src/eling/__init__.py`
- **Init** (source_file): all = ⋮---- HERMES BUDGET = 8 192 CLAUDE CLI BUDGET = 32 000 OPENCODE BUDGET = 32 000 OPENCLAW BUDGET = 24 000 OPENCLAUDE BUDGET = 24 000 ⋮---- class HarnessAdapter ABC ⋮---- name: str = "" ⋮---- @abstractmethod def read context self, project root: str Path = "." - str ⋮---- """Read the harness's memory/context file s .""" ⋮---- @abstractmethod def budget bytes self - int ⋮---- """Approximate context budget in bytes.""" ⋮---- def default schema pack self - str ⋮---- """Recommended schema pack name for this harness.""" ⋮---- def context file path self, name: str, project root: str Path = "." - Path ⋮---- root = Path project root .resolve ⋮---- candidate = parent / name ⋮---- class HermesAdap… Evidence: `src/eling/adapters/__init__.py`
- **Mcp Server** (source_file): logger = logging.getLogger name ⋮---- brain: Brain None = None brain init error: str None = None ⋮---- handshake source: str = "mcp" ⋮---- def resolve home - str None ⋮---- env = os.environ.get "ELING HOME" ⋮---- def get brain - Brain ⋮---- home = resolve home brain = Brain home=home if home else Brain brain init error = None ⋮---- brain init error = f"{type exc . name }: {exc}" ⋮---- TOOLS = ⋮---- def handle req: dict - dict None ⋮---- method = req.get "method" rid = req.get "id" params = req.get "params", {} ⋮---- def handle initialize rid: int str None, params: dict - dict ⋮---- client info = params.get "clientInfo", {} client name = client info.get "name", "unknown" client version = cli… Evidence: `src/eling/as_brain/mcp_server.py`
- **Init** (source_file): version = "0.1.0" ⋮---- all = Evidence: `src/eling/blackbox/__init__.py`
- **Init** (source_file): def detect project key cwd: str None = None - str ⋮---- result = subprocess.run ⋮---- def detect host agent - str ⋮---- def safe truncate text: str, max chars: int = 200 - str ⋮---- def load jsonl lines path: str Path - list dict str, Any ⋮---- """Load all JSON objects from a JSONL file.""" ⋮---- lines = ⋮---- line = line.strip Evidence: `src/eling/blackbox/adapters/__init__.py`
- **Block until interrupted** (source_file): def build summary from events events ⋮---- def cmd watch args: argparse.Namespace - None ⋮---- host = args.host command = args.command ⋮---- store = BlackboxStore ⋮---- adapter = ZeroAdapter on event=lambda ev: store.ingest ev cmd parts = ⋮---- Block until interrupted ⋮---- summary = adapter.get summary ⋮---- adapter = HermesAdapter ⋮---- sessions = adapter.list sessions limit=10 ⋮---- def cmd ingest args: argparse.Namespace - None ⋮---- """Ingest telemetry from a file or Hermes session.""" ⋮---- adapter = ZeroAdapter events = adapter.ingest jsonl args.zero jsonl, run id=args.run id or "" ⋮---- summary = build summary from zero events events ⋮---- events = adapter.ingest session args.hermes… Evidence: `src/eling/blackbox/cli.py`
- **Auto-score if session end** (source_file): logger = logging.getLogger name ⋮---- store: BlackboxStore None = None zero adapter: ZeroAdapter None = None hermes adapter: HermesAdapter None = None ⋮---- def get store - BlackboxStore ⋮---- db path = os.environ.get "ELING BLACKBOX DB" store = BlackboxStore db path=db path if db path else BlackboxStore ⋮---- TOOLS = ⋮---- def handle tool call rid: int str None, params: dict - dict ⋮---- tool name = params.get "name" args = dict params.get "arguments", {} store = get store ⋮---- def ok data: Any - dict ⋮---- text = json.dumps data, default=str, indent=2 ⋮---- text = json.dumps ⋮---- def err msg: str - dict ⋮---- host = args.get "host", "" cmd str = args.get "command", "" cwd = args.get "cw… Evidence: `src/eling/blackbox/mcp_server.py`
- **Layers** (source_file): logger = logging.getLogger name ⋮---- RRF K = 60 ⋮---- NOTION PAGES: dict str, tuple str, list str = { ⋮---- DEFAULT NOTION PAGE = "📋 Task Logs" ⋮---- def eling home - Path ⋮---- env = os.environ.get "ELING HOME" ⋮---- hermes home = os.environ.get "HERMES HOME" ⋮---- def detect notion category content: str, category hint: str = "" - str ⋮---- """Auto-detect Notion child page category from content + optional hint. Explicit hint wins over detection. """ ⋮---- class Brain ⋮---- Layers ⋮---- @property def task logs id self - str None ⋮---- def fire hook self, hook name: str, ctx: Any - list Any ⋮---- def ensure child page self, title: str - str None ⋮---- parent = self.notion.parent page id ⋮--… Evidence: `src/eling/brain.py`
- **── blackbox subcommand Layer 2 telemetry ──** (source_file): def main ⋮---- rest argv = a for a in sys.argv 1: if a != "--tui" ⋮---- parser = argparse.ArgumentParser ⋮---- sub = parser.add subparsers dest="cmd", required=False ⋮---- p rem = sub.add parser "remember", help="Store content" ⋮---- p rec = sub.add parser "recall", help="Search across all layers" ⋮---- p probe = sub.add parser "probe", help="Get facts about an entity" ⋮---- p reason = sub.add parser "reason", help="Compositional entity query" ⋮---- p reflect = sub.add parser "reflect", help="Promote fact id to Notion" ⋮---- p linked = sub.add parser "linked-facts", help="Get facts linked to a fact id" ⋮---- p evolve = sub.add parser ⋮---- p snap = sub.add parser ⋮---- p roll = sub.add pars… Evidence: `src/eling/cli.py`
- **Layer 4: disk config file if home is set by now, read from there** (source_file): DEFAULTS: dict str, Any = { ⋮---- ENV MAP: dict str, str = { ⋮---- TYPE MAP: dict str, type = { ⋮---- SCHEMA PACKS: dict str, dict str, Any = { ⋮---- def resolve schema pack pack name: str - dict str, Any ⋮---- base = deepcopy SCHEMA PACKS.get "default", {} ⋮---- overrides = SCHEMA PACKS.get pack name, {} ⋮---- existing = base.get key, ⋮---- def categories for pack pack name: str - list str ⋮---- def config path home: str None = None - Path ⋮---- def read json path: Path - dict ⋮---- def write json path: Path, data: dict - None ⋮---- def cast key: str, value: Any - Any ⋮---- t = TYPE MAP key ⋮---- def resolve env key: str - str None ⋮---- env name = ENV MAP.get key ⋮---- def hermes home - s… Evidence: `src/eling/config.py`
- **Init** (source_file): all = "ContinuumStore", "run stdio" Evidence: `src/eling/continuum/__init__.py`
- **Mcp Server** (source_file): logger = logging.getLogger name ⋮---- store: ContinuumStore None = None store init error: str None = None handshake source: str = "mcp" ⋮---- def resolve db - str None ⋮---- explicit = os.environ.get "ELING CONTINUUM DB" ⋮---- env = os.environ.get "ELING HOME" ⋮---- hermes = os.environ.get "HERMES HOME" ⋮---- def get store - ContinuumStore ⋮---- db = resolve db store = ContinuumStore db path=db if db else ContinuumStore store init error = None ⋮---- store init error = f"{type exc . name }: {exc}" ⋮---- TOOLS = ⋮---- def handle req: dict - dict None ⋮---- method = req.get "method" rid = req.get "id" params = req.get "params", {} ⋮---- except Exception as e: noqa: BLE001 ⋮---- def handle init… Evidence: `src/eling/continuum/mcp_server.py`
- **Decay** (source_file): DEFAULT DECAY RATE = 0.1 ⋮---- ACTIVE THRESHOLD = 0.5 ⋮---- DORMANT THRESHOLD = 0.2 ⋮---- CLEARED RECOVERY DAYS = 30 ⋮---- READ BOOST = 0.05 ⋮---- DECISION BOOST = 0.10 ⋮---- s = strength math.exp -decay rate days ⋮---- def compute lifecycle strength: float - str Evidence: `src/eling/decay.py`
- **Hermes tool registry — set by register** (source_file): logger = logging.getLogger name ⋮---- FACT REMEMBER SCHEMA = { ⋮---- FACT RECALL SCHEMA = { ⋮---- FACT FORGET SCHEMA = { ⋮---- FACT PROBE SCHEMA = { ⋮---- FACT STATS SCHEMA = { ⋮---- FACT EVOLVE SCHEMA = { ⋮---- FACT ENTITY NEIGHBORS SCHEMA = { ⋮---- class FactMemoryProvider ⋮---- Hermes tool registry — set by register ⋮---- ------------------------------------------------------------------ Lazy FactsLayer ⋮---- @property def facts self - FactsLayer ⋮---- """Lazy-initialized FactsLayer instance.""" ⋮---- Core API ⋮---- pp = self. privacy.process content ⋮---- clean = pp "clean" fact id = self.facts.add clean, category=category, tags=tags, source=source ⋮---- cat = category or None ⋮---- def… Evidence: `src/eling/fact_memory_provider.py`
- **Build query from tool name + stringified args** (source_file): logger = logging.getLogger name ⋮---- HOOK SESSION START = "session start" HOOK PRE USER MESSAGE = "pre user message" HOOK POST USER MESSAGE = "post user message" HOOK PRE TOOL USE = "pre tool use" HOOK POST TOOL USE = "post tool use" HOOK POST ASSISTANT MESSAGE = "post assistant message" HOOK DECISION MADE = "decision made" HOOK FILE EDIT = "file edit" HOOK ERROR OCCURRED = "error occurred" HOOK COMPACTION = "compaction" HOOK SESSION END = "session end" HOOK IDLE 30MIN = "idle 30min" ⋮---- HOOK VERIFY REQUEST = "verify request" ⋮---- HOOK SYNC START = "sync start" HOOK SYNC COMPLETE = "sync complete" HOOK SYNC ERROR = "sync error" ⋮---- ALL HOOKS = ⋮---- HookHandler = Callable str, dict ,… Evidence: `src/eling/hooks.py`
- **Builtin** (source_file): class BuiltinLayer ⋮---- hermes home = os.environ.get "HERMES HOME", "~/.hermes" memory path = Path hermes home .expanduser / "MEMORY.md" ⋮---- user path = Path hermes home .expanduser / "USER.md" ⋮---- @property def available self - bool ⋮---- def read memory self - str ⋮---- def read user self - str ⋮---- def search self, query: str - list dict ⋮---- """Simple substring grep across both files.""" out = q = query.lower ⋮---- content = path.read text encoding="utf-8", errors="replace" Evidence: `src/eling/layers/builtin.py`
- **Code** (source_file): class CodeLayer ⋮---- cache path = self.project path / ".eling" / "code index.json" ⋮---- def init index self - None ⋮---- @property def available self - bool ⋮---- @property def symbol count self - int ⋮---- @property def file count self - int ⋮---- def search self, query: str, max files: int = 12 - list dict ⋮---- def explore self, query: str, max files: int = 12 - dict ⋮---- def reindex self, file path: str Path - bool ⋮---- def build index self - int Evidence: `src/eling/layers/code.py`
- **── Temporal / Date parsing ───────────────────────────────────────────────** (source_file): logger = logging.getLogger name ⋮---- SCHEMA = """ ⋮---- HELPFUL DELTA = 0.05 UNHELPFUL DELTA = -0.10 ⋮---- RE CAPITALIZED = re.compile r"\b A-Z a-z + ?:\s+ A-Z a-z + + \b" RE DOUBLE QUOTE = re.compile r'" ^" + "' RE SINGLE QUOTE = re.compile r"' ^' + '" RE WIKI LINK = re.compile r"\ \ ^\ + \ \ " ⋮---- ── Temporal / Date parsing ─────────────────────────────────────────────── RELATIVE PATTERNS: list tuple re.Pattern, str, int = ⋮---- Numbered relative: "last X days/weeks/months" ⋮---- """Patterns for relative time expressions. Each entry: pattern, unit, offset direction .""" ⋮---- ABSOLUTE DATE PATTERNS: list tuple re.Pattern, str = """Absolute date patterns: ISO 8601, US-style, named-month… Evidence: `src/eling/layers/facts.py`
- **Kb** (source_file): logger = logging.getLogger name ⋮---- SCHEMA = """ Evidence: `src/eling/layers/kb.py`
- **fall through to blocks walk on empty/failed full retrieval** (source_file): HAS HTTPX: bool None = None ⋮---- logger = logging.getLogger name ⋮---- NOTION API = "https://api.notion.com/v1" NOTION VERSION = "2022-06-28" ⋮---- def require httpx ⋮---- mod = importlib.import module "httpx" HAS HTTPX = True ⋮---- HAS HTTPX = False ⋮---- class NotionLayer ⋮---- @property def available self - bool ⋮---- def has httpx self - bool ⋮---- def get client self - t.Any ⋮---- httpx = require httpx ⋮---- def search self, query: str, limit: int = 10 - list dict ⋮---- r = self. get client .post ⋮---- data = r.json results = ⋮---- title = self. extract title page ⋮---- def get page markdown self, page id: str, prefer full: bool = True - str ⋮---- full = self.get page full markdown pa… Evidence: `src/eling/layers/notion.py`
- **Init** (source_file): all = "run stdio" Evidence: `src/eling/markdownify/__init__.py`
- **── Tool definitions ─────────────────────────────────────────────────────────** (source_file): logger = logging.getLogger name ⋮---- md instance = None ⋮---- def get markitdown ⋮---- md instance = MarkItDown ⋮---- md instance = False ⋮---- def resolve allowed - list Path None ⋮---- raw = os.environ.get "MD ALLOWED PATHS" or os.environ.get "MD SHARE DIR" ⋮---- sep = ";" if os.name == "nt" else ":" ⋮---- def check path filepath: str, allowed: list Path None - Path ⋮---- p = Path filepath .expanduser .resolve ⋮---- ── Tool definitions ───────────────────────────────────────────────────────── ⋮---- TOOLS = ⋮---- def handle req: dict - dict None ⋮---- method = req.get "method" rid = req.get "id" params = req.get "params", {} ⋮---- def handle initialize rid: int str None, params: dict - di… Evidence: `src/eling/markdownify/mcp_server.py`
- **Mcp Server** (source_file): logger = logging.getLogger name ⋮---- notion: NotionLayer None = None ⋮---- def get notion - NotionLayer ⋮---- notion = NotionLayer ⋮---- TOOLS = ⋮---- def handle req: dict - dict None ⋮---- method = req.get "method" rid = req.get "id" params = req.get "params", {} ⋮---- def handle initialize rid: int str None, params: dict - dict ⋮---- client info = params.get "clientInfo", {} client name = client info.get "name", "unknown" client version = client info.get "version", "?" ⋮---- def handle tool call rid: int str None, params: dict - dict ⋮---- tool name = params.get "name" args = dict params.get "arguments", {} notion = get notion ⋮---- def ok data: Any - dict ⋮---- text = json.dumps data, d… Evidence: `src/eling/mcp_server.py`
- **Permissions** (source_file): ACCESS LEVELS = "write", "read", "none" KNOWN LAYERS = "facts", "kb", "code", "notion", "hrr" ⋮---- def default path - Path ⋮---- eling home = Path ⋮---- def load permissions path: str Path None = None - dict str, Any ⋮---- p = Path path if path else default path ⋮---- data = json.loads p.read text encoding="utf-8" ⋮---- def canonical access raw: str - str ⋮---- access = raw.strip .lower ⋮---- perms = load permissions ⋮---- sources = perms.get "sources", {} ⋮---- source cfg = sources.get source ⋮---- access = source cfg.get layer, "write" access = canonical access access ⋮---- def describe permissions perms: dict str, Any None = None - list dict ⋮---- rows: list dict = ⋮---- access = canoni… Evidence: `src/eling/permissions.py`
- **Patterns that should produce a warning but not full redaction e.g., env file lines** (source_file): SECRET PATTERNS: list tuple str, Pattern str = ⋮---- Patterns that should produce a warning but not full redaction e.g., env file lines SENSITIVE PATTERNS: list tuple str, Pattern str = ⋮---- def strip secrets text: str - tuple str, list str ⋮---- """Strip known secret patterns from text. Returns: sanitized text, redacted kinds — cleaned text and list of what was redacted. """ sanitized = text redacted: set str = set ⋮---- sanitized = pattern.sub f" REDACTED:{name} ", sanitized ⋮---- Also warn about sensitive-but-not-destructive patterns ⋮---- def redact kinds kinds: list str - str ⋮---- """Pretty-print redacted kinds for logging.""" ⋮---- ---------------------------------------------------… Evidence: `src/eling/privacy.py`
- **── Core ops ──────────────────────────────────────────────────────────────────** (source_file): logger = logging.getLogger name ⋮---- SNAPSHOT KEEP = 5 ⋮---- def snapshot dir parent: Path - Path ⋮---- d = parent / "snapshots" ⋮---- snapshot counter = itertools.count ⋮---- def snapshot id - str ⋮---- ── Core ops ────────────────────────────────────────────────────────────────── ⋮---- """Snapshot the facts database. Returns snapshot metadata. Parameters ---------- db path : str or Path Path to the current facts.db. reason : str Why the snapshot was taken e.g. "pre decay", "pre contradiction resolve" . keep : int Max snapshots to keep. Oldest are pruned after creating the new one. Returns ------- dict with keys: snapshot id, path, reason, size bytes, fact count, timestamp. Return all ava… Evidence: `src/eling/snapshot.py`
- **Cli** (source_file): repo root = os.path.dirname os.path.dirname os.path.dirname os.path.abspath file ⋮---- all = "main" Evidence: `src/eling_agent/cli.py`
- **Core Architecture** (documentation): - Agent-Blackbox Nous Research / Taewoo Park https://github.com/joint79 — Blackbox Layer 2 flight recorder is a port of the Agent-Blackbox concept: 11-metric context-efficiency scoring, SQLite event store, causal timeline builder, optimization suggestion engine, and per-archetype baselines. MIT https://github.com/nousresearch/agent-blackbox - Continuum multi-agent orchestration — worktree isolation, PLOT protocol, and agent dispatch registry inspired by continuum https://github.com/pouyahasanamreji/continuum by Pouya Hasanamreji https://github.com/pouyahasanamreji . - Spec-kit integration — spec-driven development artifacts from spec-kit https://github.com/github/spec-kit by GitHub. MIT htt… Evidence: `CREDITS.library.md`
- **Credits** (documentation): Eling Agent was created by PatrickNoFilter @PatrickNoFilter https://github.com/PatrickNoFilter . Evidence: `CREDITS.md`
- **.gitignore** (source_file): pycache / .pyc .db .db-shm .db-wal config.json .env .venv venv/ .egg-info/ dist/ build/ .git/ .eling/ .agent-version Evidence: `.gitignore`
- **numpy is optional — needed only for HRR.** (source_file): build-system requires = "setuptools =68.0" build-backend = "setuptools.build meta" Evidence: `pyproject.library.toml`

## Rules the Host AI Must Follow

- **Treat this asset as pre-work context, not a runtime environment.**: The AI Context Pack contains only an evidence-backed understanding of the project, not the project's executable state. Evidence: `README.md`, `src/eling/zero_plugin/SKILL.md`, `docs/architecture.md`
- **When answering the user, distinguish what can be previewed from what can only be verified after install.**: The consumer value of the pre-install experience comes from reducing bad installs and misjudgments, not from pretending to be a real run. Evidence: `README.md`, `src/eling/zero_plugin/SKILL.md`, `docs/architecture.md`

## Questions the User Should Answer First

- Which host AI or local environment do you plan to use it in?
- Do you just want to experience the workflow first, or are you ready to actually install?
- What matters most to you: install cost, output quality, or conflicts with your existing rules?

## Acceptance Checks

- Every capability claim can be traced back to a file path in evidence_refs.
- AI_CONTEXT_PACK.md does not package previews as a real run.
- The user can understand who it fits, what it can do, how to start, and the risk boundaries within 3 minutes.

---

## Doramagic Context Augmentation

The following sections strengthen the repository context for a host AI. Human Manual data is a reading route, and pitfall notes become operating constraints.

## Human Manual Outline

Usage rule: this is only a reading route and salience signal, not factual authority. Concrete claims must still return to repo evidence or Claim Graph.

Host AI hard rules:
- Do not treat page titles, section order, summaries, or importance values as factual project evidence.
- When explaining the Human Manual outline, state that it is only a reading route or salience signal.
- Capability, installation, compatibility, runtime state, and risk claims must cite repo evidence, source paths, or Claim Graph.

- **Introduction and Quick Start**: importance `high`
  - source_paths: README.md, agent.py, requirements.txt, pyproject.toml
- **System Architecture and Component Map**: importance `high`
  - source_paths: agent.py, provider.py, mcp_client.py, memory.py, skills.py
- **Configuration, Setup Wizard, and Themes**: importance `high`
  - source_paths: config.example.json, src/eling/cli.py, src/eling/config.py, docs/configuration.md
- **Terminal UI (TUI), Spinner, and Theming**: importance `high`
  - source_paths: tui.py, src/eling/config.py
- **Memory System: 8-Layer Second Brain**: importance `high`
  - source_paths: memory.py, textsim.py, src/eling/brain.py, src/eling/layers/builtin.py, src/eling/layers/code.py
- **Skill Library and Self-Learning**: importance `high`
  - source_paths: skills.py, src/eling/zero_plugin/SKILL.md, src/eling/decay.py, src/eling/snapshot.py
- **Plugin System, MCP Servers, and the @tool Decorator**: importance `high`
  - source_paths: plugins/__init__.py, plugins/shell_plugin.py, plugins/eling_integration.py, mcp_client.py, src/eling/mcp_server.py
- **Tool Orchestration: Workspace, Auto-pytest, and Auto-fix Lint**: importance `high`
  - source_paths: agent.py, plugins/shell_plugin.py, plugins/eling_integration.py, src/eling/hooks.py, src/eling/permissions.py

## Repo Inspection Evidence

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `fe8590b7751e20fcb3bc0764e61889164ddd0ad3`
- inspected_files: `README.md`, `pyproject.toml`, `requirements.txt`, `docs/architecture.md`, `docs/configuration.md`, `docs/index.md`, `docs/mcp.md`, `docs/memory.md`, `docs/plugins.md`, `src/eling/__init__.py`, `src/eling/adapters/__init__.py`, `src/eling/as_brain/__init__.py`, `src/eling/as_brain/mcp_server.py`, `src/eling/blackbox/__init__.py`, `src/eling/blackbox/adapters/__init__.py`, `src/eling/blackbox/adapters/common.py`, `src/eling/blackbox/adapters/hermes.py`, `src/eling/blackbox/adapters/zero.py`, `src/eling/blackbox/cli.py`, `src/eling/blackbox/core.py`

Host AI hard rules:
- Without repo_clone_verified=true, do not claim that the source code has been read.
- Without repo_inspection_verified=true, do not write README, docs, or package-file conclusions as facts.
- Without quick_start_verified=true, do not claim that the Quick Start path has run successfully.

## Doramagic Pitfall Constraints

These rules come from Doramagic discovery, validation, or compilation findings. The host AI must treat them as operating constraints, not background notes.

### Constraint 1: Capability evidence risk requires verification

- Trigger: README/documentation is current enough for a first validation pass.
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: capability.assumptions | https://github.com/PatrickNoFilter/eling-agent
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 2: Security or permission risk requires verification

- Trigger: no_demo
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: downstream_validation.risk_items | https://github.com/PatrickNoFilter/eling-agent
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 3: Security or permission risk requires verification

- Trigger: no_demo
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: risks.scoring_risks | https://github.com/PatrickNoFilter/eling-agent
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 4: Maintenance risk requires verification

- Trigger: issue_or_pr_quality=unknown。
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: evidence.maintainer_signals | https://github.com/PatrickNoFilter/eling-agent
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 5: Maintenance risk requires verification

- Trigger: release_recency=unknown。
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: evidence.maintainer_signals | https://github.com/PatrickNoFilter/eling-agent
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.